D 1380
OULU 2016
UNIVERSITY OF OUL U P.O. Box 8000 FI-90014 UNIVERSITY OF OULU FINLA ND
U N I V E R S I TAT I S
O U L U E N S I S
ACTA
A C TA
D 1380
ACTA
U N I V E R S I T AT I S O U L U E N S I S
Pekka Pinola
University Lecturer Santeri Palviainen
Postdoctoral research fellow Sanna Taskila
Professor Olli Vuolteenaho
University Lecturer Veli-Matti Ulvinen
Director Sinikka Eskelinen
Pekka Pinola
Professor Esa Hohtola
HYPERANDROGENISM,
MENSTRUAL
IRREGULARITIES AND
POLYCYSTIC OVARY
SYNDROME
IMPACT ON FEMALE REPRODUCTIVE AND
METABOLIC HEALTH FROM EARLY ADULTHOOD
UNTIL MENOPAUSE
Professor Jari Juga
University Lecturer Anu Soikkeli
Professor Olli Vuolteenaho
Publications Editor Kirsti Nurkkala
ISBN 978-952-62-1309-5 (Paperback)
ISBN 978-952-62-1310-1 (PDF)
ISSN 0355-3221 (Print)
ISSN 1796-2234 (Online)
UNIVERSITY OF OULU GRADUATE SCHOOL;
UNIVERSITY OF OULU, FACULTY OF MEDICINE;
MEDICAL RESEARCH CENTER OULU;
OULU UNIVERSITY HOSPITAL;
NATIONAL INSTITUTE FOR HEALTH AND WELFARE
D
MEDICA
ACTA UNIVERSITATIS OULUENSIS
D Medica 1380
PEKKA PINOLA
HYPERANDROGENISM, MENSTRUAL
IRREGULARITIES AND POLYCYSTIC
OVARY SYNDROME
Impact on female reproductive and metabolic health
from early adulthood until menopause
Academic dissertation to be presented with the assent of
the Doctoral Training Committee of Health and
Biosciences of the University of Oulu for public defence in
Auditorium 4 of Oulu University Hospital, on 9
September 2016, at 12 noon
U N I VE R S I T Y O F O U L U , O U L U 2 0 1 6
Copyright © 2016
Acta Univ. Oul. D 1380, 2016
Supervised by
Docent Laure Morin-Papunen
Reviewed by
Docent Mervi Halttunen-Nieminen
Professor Leo Niskanen
Opponent
Professor Okan Bülent Yıldız
ISBN 978-952-62-1309-5 (Paperback)
ISBN 978-952-62-1310-1 (PDF)
ISSN 0355-3221 (Printed)
ISSN 1796-2234 (Online)
Cover Design
Raimo Ahonen
JUVENES PRINT
TAMPERE 2016
Pinola, Pekka, Hyperandrogenism, menstrual irregularities and polycystic ovary
syndrome. Impact on female reproductive and metabolic health from early
adulthood until menopause
University of Oulu Graduate School; University of Oulu, Faculty of Medicine; Medical
Research Center Oulu; Oulu University Hospital; National Institute for Health and Welfare
Acta Univ. Oul. D 1380, 2016
University of Oulu, P.O. Box 8000, FI-90014 University of Oulu, Finland
Abstract
Polycystic ovary syndrome (PCOS) is the most common endocrine disorder of women of
reproductive age, affecting 5–18% of them. Menstrual irregularities, hyperandrogenemia and
obesity are key features in PCOS and they are suggested to be the most important metabolic risks
linked to PCOS, but their respective roles are still under debate. Anti-Müllerian hormone (AMH)
is involved in sexual differentiation and follicle growth and its level is increased in women with
PCOS.
The aims of this project were to clarify the significance of menstrual irregularities,
hyperandrogenemia and serum levels of AMH in adolescence as predictive factors of the
syndrome and to investigate the respective roles of obesity and hyperandrogenism as metabolic
risk factors in women with PCOS from adolescence to late adulthood.
The study populations were the Northern Finland Birth Cohort 1986 (N=3373 women) and a
Nordic population including 1553 women with PCOS and 448 controls.
At the age of 16 years, women with menstrual irregularities were more hyperandrogenic
compared with women with normal menstrual cycles. Serum AMH levels correlated positively
with those of testosterone at this age. They were higher in adolescents with menstrual irregularities
compared with those with regular cycles and in women with hirsutism or PCOS at the age of 26
years. However, AMH was not a good marker of metabolic abnormalities in adolescence or a
reliable tool to predict PCOS in later life. Androgen levels were higher in women with PCOS
throughout life compared with controls. The parameters that best predicted PCOS at all ages were
the free androgen index, and androstenedione. Women with PCOS exhibited increased abdominal
obesity, altered insulin metabolism, worse lipid profiles and higher blood pressure from early
adulthood until menopause compared with controls. The highest prevalence of metabolic
syndrome was detected in obese and hyperandrogenic women with PCOS.
In conclusion, irregular menstrual cycles, identified by a simple question at adolescence,
represent a good marker of hyperandrogenemia, later metabolic risks and development of PCOS.
Due to the persistence of hyperandrogenism and metabolic alterations, the treatment of PCOS
should be focused on prevention and treatment of these problems as early as in adolescence in
order to decrease future morbidity.
Keywords: aging, androgens, cardiovascular diseases, hyperandrogenism, metabolism,
polycystic ovary syndrome, reproductive health
Pinola, Pekka, Mieshormoniylimäärä, kuukautishäiriöt ja monirakkulainen
munasarjaoireyhtymä. Vaikutus naisen lisääntymis- ja aineenvaihduntaterveyteen
varhaisaikuisuudesta menopaussiin
Oulun yliopiston tutkijakoulu; Oulun yliopisto, Lääketieteellinen tiedekunta; MRC Oulututkimusyksikkö; Oulun yliopistollinen sairaala; Terveyden ja hyvinvoinnin laitos
Acta Univ. Oul. D 1380, 2016
Oulun yliopisto, PL 8000, 90014 Oulun yliopisto
Tiivistelmä
Monirakkulainen munasarjaoireyhtymä (polycystic ovary syndrome, PCOS) on lisääntymisikäisten naisten tavallisin (5-18%) hormonaalinen häiriö. Kuukautishäiriöt, mieshormoniylimäärä eli hyperandrogenismi ja lihavuus kuuluvat oireyhtymään oleellisesti ja niiden ajatellaan olevan tärkeimmät PCOS:aan liittyvät metaboliset riskitekijät, vaikkakin niiden tarkat roolit ovat epäselvät. Anti-Müllerian hormoni (AMH) vaikuttaa sukupuolen kehitykseen sikiöaikana sekä munarakkuloiden kypsymiseen hedelmällisessä iässä ja sen pitoisuus on suurentunut PCOS-naisilla.
Tämän väitöskirjan tavoitteena oli selvittää kuukautishäiriöiden, hyperandrogenismin ja
AMH-pitoisuuden ennustearvoa PCOS-oireyhtymälle, sekä arvioida lihavuuden ja hyperandrogenismin vaikutusta metabolisiin riskitekijöihin PCOS-naisilla läpi elämän. Tutkimusaineistoina olivat Pohjois-Suomen syntymäkohortti 1986 (N=3373 naista) sekä pohjoismaalainen yhteistyöaineisto, jossa oli 1553 PCOS-naista ja 448 kontrollia.
16-vuotiaiden kuukautishäiriöistä kärsivien naisten mieshormonipitoisuuksien todettiin olevan korkeammat kuin naisilla, joilla kuukautiskierto oli säännöllinen. AMH-tasot korreloituivat
positiivisesti testosteronin kanssa 16-vuotiaana, ja pitoisuus oli koholla naisilla, joilla todettiin
kuukautishäiriö. AMH-tasot 16-vuotiaana olivat myös korkeampia naisilla, joilla oli 26-vuotiaana PCOS tai hirsutismi. Kuitenkaan AMH-pitoisuus 16-vuotiaana ei korreloinut metabolisten
riskitekijöiden kanssa eikä se ollut luotettava parametri ennustamaan PCOS:n kehittymistä.
Mieshormonitasot olivat korkeammat PCOS-naisilla läpi elämän verrattuna kontrolleihin.
Vapaan mieshormonin indeksit ja androstendioni olivat parhaat parametrit erottamaan PCOSnaiset kontrolleista. PCOS-naisilla todettiin olevan enemmän vyötärölihavuutta, huonompi veren
rasvaprofiili ja korkeampi verenpaine varhaisaikuisuudesta menopaussiin saakka. Lihavilla
hyperandrogeenisilla naisilla todettiin suurin metabolisen oireyhtymän esiintyvyys.
Yksinkertaisella kysymyksellä selvitetyt kuukautishäiriöt nuoruusiässä todettiin olevan yhteydessä hyperandrogenismiin sekä myöhempiin metabolisiin riskeihin ja PCOS:n kehittymiseen.
Jotta pystyttäisiin vähentämään myöhempää sairastavuutta ja kuolleisuutta PCOS-naisilla, oireyhtymän hoidon tulisi keskittyä ennaltaehkäisemään ja hoitamaan lihavuutta, hyperandrogenismia ja metabolisia riskejä jo varhaisaikuisuudessa.
Asiasanat: androgeenit, hyperandrogenismi, lisääntymisterveys, metabolia, mieshormoniylimäärä, monirakkulainen munasarjaoireyhtymä, sydän- ja verisuonisairaudet
The secret of life, though, is to fall seven times
and to get up eight times.
Paulo Coelho
To my family
8
Acknowledgements
The present study was carried out at the Department of Obstetrics and Gynecology,
University of Oulu and the Medical Research Center Oulu, Oulu University
Hospital, in collaboration with the National Institute for Health and Welfare. The
first two papers were based on data from the Northern Finland Birth Cohort 1986
and without the contribution of two magnificent ladies, Professor Marjo RiittaJärvelin and Professor Anna-Liisa Hartikainen, the project would not have even
started. Their commitment to data collection and supporting young students is
essential for research in Northern Finland.
The present study was also dependent on great Nordic collaborative work and
I wish to thank our highly important collaborators, Professor Eszter Vanky,
Professor Inger Sundström-Poromaa, Docent Elisabet Stener-Victorin, Professor
Angelica Lindén Hirschberg, Docent Pernille Ravn, Professor Marianne Skovsager
Andersen, Dr. Dorte Glintborg, M.D., Ph.D. and Dr. Jan Roar Mellembakken, M.D.,
Ph.D. Thank you for supporting young students in Finland.
I express my deepest gratitude and respect to my “mother-in-science”, Docent
Laure Morin-Papunen. Her guidance and support has been invaluable ever since I
first stepped into her office as a first-year medical student in 2010. Her enthusiasm
has helped me over the obstacles during these past few years and she has giving me
inspiration and strength when they seemed to have gone.
I am very grateful to Professor Juha Tapanainen, who has always been there
when needed, giving precious guidance, both in science and in life. In the same
way, Docent Terhi Piltonen has been there for a young student, always helping with
difficulties, sharing her great knowledge and creating an inspiring scientific
environment.
I wish to thank Professor Hannu Martikainen and Docent Anneli Pouta for
supporting the early steps on my scientific road. Special thanks go to Professor
Stephen Franks who has been kind, understanding and supportive throughout this
whole project.
I wish to thank the reviewers of this thesis, Docent Mervi Halttunen-Nieminen
and Professor Leo Niskanen. I express my sincerest gratitude to my follow-up
colleagues Docent Marja Vääräsmäki and Dr. Eila Suvanto, M.D., Ph.D. for their
support and understanding. Even though I did not succeed in organizing meetings
as well as I could have, they have guided me through the project at informal times
in the corridors and at coffee breaks.
9
I thank Professor Aimo Ruokonen for his wise words and great help with the
clinical chemistry in the present study. I also thank Docent Katri Puukka and Mari
Ulmanen, M.Sc. for their valuable work in the laboratory. I also express my
gratitude to Aini Bloigu and Risto Bloigu for their guidance in statistics. I thank
Nicholas Bolton for linguistic revision throughout this whole project.
I owe a great deal to the younger scientists in our group, Johanna Puurunen,
Sammeli West, Meri-Maija Ollila, Salla Karjula, Marika Kangasniemi, Saara
Vuontisjärvi, Anni Rantala, Sanna Koskela, Masuma Khatun and Shilpa Lingaiah.
The youthful atmosphere is definitely one of the most important supporting
structures in our study group, both in everyday work and relaxation. In science,
Fanni Päkkilä has been the greatest peer support. Side by side we started our work
together and now we have grown to enjoy a lifetime of friendship; thank you Fanni
for everything.
I thank all my friends for their important support and never-ending kindness.
All the events, trips and late-night conversations have given me inspiration and
strength to carry on, even in the most troubled times. I thank you for being in my
life.
I have received valuable support from all the places that I have worked in. I
thank all the lovely and friendly people in Kokkola Central Hospital, in Ylivieska
Health Center and in the Gynecology Department of Oulu University Hospital.
My deepest respect goes to my parents Eila and Eino Pinola and to my siblings
Piia-Maria Vähäkangas, Kirsi Nurkkala and Veikko Nurkkala. In my early
childhood you were all there raising me, helping me learn to read, to work hard and
above all to appreciate the things I have been given. Not a single wish I ever had
was denied by you. Thank you for having faith in me and supporting me to chase
my dreams.
Above all, the greatest support came from my partner Samuli. Even though he
has seen all the worst parts of scientific work, and probably me too, he has gone
through it all, never leaving my side. His encouragement and reality checks every
now and then helped me to balance work and leisure. Thank you Samuli.
This study was financially enabled by Oulu University Hospital, the Northern
Finland Health Care Support Foundation, the Finnish Medical Society and the
Finnish Medical Foundation, to whom I am grateful.
Oulu, July 2016
10
Pekka Pinola
Abbreviations
A4
ACTH
AE-PCOS
AFC
AMH
ANCOVA
ANOVA
BMI
BP
cFT
CRP
CVD
DHEA
DHEAS
DHT
e.g.
F&G
FAI
FSH
FT
GC-MS
GnRH
HA
HA-PCOS
hCG
HDL
HOMA-IR
HOMA-S
HPO
hs-CRP
i.e.
IGT
IR
LC-MS/MS
LDL
Androstenedione
Adrenocorticotropic hormone
Androgen Excess and Polycystic Ovary Syndrome
Antral follicle count
Anti-Müllerian hormone
Analysis of covariance
Analysis of variance
Body mass index
Blood pressure
Calculated free testosterone
C-reactive protein
Cardiovascular disease
Dehydroepiandrosterone
Dehydroepiandrosterone sulfate
Dihydrotestosterone
exempli gratia, for example
Ferriman and Gallwey
Free androgen index
Follicle-stimulating hormone
Free testosterone
Gas chromatography-mass spectrometry
Gonadotropin-releasing hormone
Hyperandrogenism
Hyperandrogenic (woman) with PCOS
Human chorionic gonadotropin
High-density lipoprotein
Homeostasis model assessment of insulin resistance
Homeostasis model assessment of insulin sensitivity
Hypothalamus–pituitary–ovary
High-sensitivity C-reactive protein
id est, that is
Impaired glucose tolerance
Insulin resistance
Liquid chromatography-tandem mass spectrometry
Low-density lipoprotein
11
LH
Luteinizing hormone
MetS
Metabolic syndrome
NAFLD
Non-alcoholic fatty liver disease
NA-PCOS Normoandrogenic (woman) with PCOS
NFBC-1986 Northern Finland Birth Cohort 1986
NIH
National Institutes of Health
OA
Oligo-amenorrhea
OGTT
Oral glucose tolerance test
OR
Odds ratio
OSA
Obstructive sleep apnea
P450c17
17-hydroxylase and 17,20 lyase activity
P450scc
Side-chain cleavage enzyme
PCOM
Polycystic ovary morphology
PCOS
Polycystic ovary syndrome
RIA
Radioimmunoassay
ROC
Receiver operating characteristic
SD
Standard deviation
SES
Socioeconomic status
SHBG
Sex hormone-binding globulin
StAR
Steroidogenic acute regulatory protein
T
Testosterone
T2DM
Type II diabetes mellitus
TG
Triglyceride
WHR
waist-to-hip ratio
12
List of original publications
This thesis is based on the following publications, which are referred to throughout
the text by their Roman numerals:
I
Pinola P, Lashen H, Bloigu A, Puukka K, Ulmanen M, Ruokonen A, Martikainen H,
Pouta A, Franks S, Hartikainen AL, Järvelin MR & Morin-Papunen L (2012) Menstrual
disorders in adolescence: a marker for hyperandrogenaemia and increased metabolic
risks in later life? Finnish general population-based birth cohort study. Hum Reprod
27(11): 3279-3286.
II Pinola P, Morin-Papunen LC, Bloigu A, Puukka K, Ruokonen A, Järvelin MR, Franks
S, Tapanainen JS & Lashen H (2014) Anti-Müllerian hormone: correlation with
testosterone and oligo- or amenorrhoea in female adolescence in a population-based
cohort study. Hum Reprod 29(10): 2317-2325.
III Pinola P, Piltonen TT, Puurunen J, Vanky E, Sundström-Poromaa I, Stener-Victorin E,
Ruokonen A, Puukka K, Tapanainen JS & Morin-Papunen LC (2015) Androgen Profile
Through Life in Women With Polycystic Ovary Syndrome: A Nordic Multicenter
Collaboration Study. J Clin Endocrinol Metab 100(9): 3400-3407.
IV Pinola P, Puukka K, Piltonen TT, Puurunen J, Vanky E, Sundström-Poromaa I, StenerVictorin E, Lindén Hirschberg A, Ravn P, Skovsager Andersen M, Glintborg D, Roar
Mellembakken J, Ruokonen A, Tapanainen JS & Morin-Papunen LC (2016) Both
normo- and hyperandrogenic women with polycystic ovary syndrome exhibit an
adverse metabolic profile from early adulthood to menopause: a Nordic multicenter
collaboration study. Manuscript.
13
14
Contents
Abstract
Tiivistelmä
Acknowledgements
9
Abbreviations
11
List of original publications
13
Contents
15
1 Introduction
19
2 Review of the literature
21
2.1 Androgen secretion in women ................................................................ 21
2.1.1 Ovarian androgen secretion .......................................................... 24
2.1.2 Adrenal androgen secretion .......................................................... 24
2.1.3 Peripheral tissues .......................................................................... 25
2.2 Hyperandrogenism in women ................................................................. 25
2.2.1 Evaluation of hyperandrogenism in women ................................. 26
2.3 Anti-Müllerian hormone ......................................................................... 31
2.4 Polycystic ovary syndrome ..................................................................... 33
2.4.1 Definition and prevalence............................................................. 33
2.4.2 Hyperandrogenism in PCOS ........................................................ 34
2.4.3 Menstrual irregularities in PCOS ................................................. 35
2.4.4 PCO morphology at ultrasonography in PCOS ............................ 36
2.4.5 Other clinical features of PCOS ................................................... 36
2.4.6 Metabolic risks in PCOS .............................................................. 39
2.4.7 Non-alcoholic fatty liver disease (NAFLD) and
obstructive sleep apnea (OSA) in PCOS ...................................... 41
2.5 PCOS from adolescence to postmenopause ............................................ 42
2.5.1 Prepubertal origins of PCOS ........................................................ 43
2.5.2 PCOS in adolescence.................................................................... 43
2.5.3 PCOS at reproductive age ............................................................ 44
2.5.4 PCOS in premenopause and menopause ...................................... 45
3 Purpose of the present study
47
4 Subjects and methods
49
4.1 Study populations and respective methods ............................................. 49
4.1.1 Northern Finland Birth Cohort 1986 (Studies I and II) ................ 49
4.1.2 Nordic multicenter data (Studies III and IV) ................................ 51
4.1.3 Methods ........................................................................................ 52
15
4.2 Data analysis and statistical methods ...................................................... 54
5 Results
59
5.1 Menstrual disorders and hyperandrogenemia in adolescence and
early adulthood (Studies I and II) ............................................................ 59
5.1.1 Menstrual disorders ...................................................................... 59
5.1.2 Effect of hyperandrogenemia ....................................................... 60
5.1.3 Effect of weight ............................................................................ 60
5.1.4 Anti-Müllerian hormone ............................................................... 61
5.2 Endocrine features of PCOS throughout life (Study III) ......................... 62
5.2.1 Testosterone, FAI and cFT ............................................................ 63
5.2.2 Adrenal androgens ........................................................................ 65
5.2.3 Sex hormone-binding globulin ..................................................... 66
5.2.4 Menstrual pattern .......................................................................... 66
5.2.5 Evaluation of the risk of PCOS .................................................... 66
5.3 Metabolic features of PCOS throughout life (Study IV) ......................... 67
5.3.1 Comparison of women with PCOS and control women in
the whole population .................................................................... 67
5.3.2 Comparisons between HA-PCOS, NA-PCOS and control
women and the different age groups ............................................. 70
5.3.3 Body mass index and waist circumference ................................... 70
5.3.4 Glucose and insulin metabolism ................................................... 71
5.3.5 Lipids ............................................................................................ 72
5.3.6 Blood pressure .............................................................................. 74
5.3.7 High-sensitivity C-reactive protein............................................... 75
5.3.8 Prevalence of metabolic syndrome ............................................... 75
6 Discussion
77
6.1 Early prediction of PCOS........................................................................ 77
6.1.1 Menstrual irregularities................................................................. 77
6.1.2 Anti-Müllerian hormone ............................................................... 79
6.2 Lifelong hormonal and metabolic parameters ......................................... 80
6.2.1 Androgen parameters in PCOS across life ................................... 80
6.2.2 Metabolic risk factors in PCOS across life ................................... 81
6.3 Strengths and limitations of the studies................................................... 84
6.3.1 Studies I and II (NFBC-1986) ...................................................... 84
6.3.2 Studies III and IV (Nordic population) ......................................... 85
7 Conclusions and future plans
87
7.1.1 Future plans .................................................................................. 88
16
References
Original publications
89
107
17
18
1
Introduction
Polycystic ovary syndrome (PCOS) is the most common gynecological
endocrinopathy, affecting women of fertile age, and the most common cause of
infertility in women. The prevalence of PCOS varies from 6% to 18% depending
on the diagnostic criteria used (March et al. 2010). According to the Rotterdam
consensus (Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop
group 2004) diagnosis of the syndrome requires at least two of the following
features: polycystic ovaries detected by ultrasonography, oligo- or anovulation
and/or either clinical (hirsutism) or biochemical (high serum testosterone or free
androgen index [FAI]) hyperandrogenism.
PCOS usually has its onset early in life. Typically, the cardinal features of the
syndrome are most severe during the reproductive years and the severity of
hyperandrogenism and menstrual disorders decrease premenopausally (Welt &
Carmina 2013). In addition, women with PCOS may present typical metabolic
abnormalities such as insulin resistance (IR) and visceral obesity at a young age.
Long-term exposure to these abnormalities throughout fertile life may exacerbate
the adverse effects and expose these women to higher risks of metabolic syndrome
(MetS), cardiovascular diseases (CVDs) and type II diabetes mellitus (T2DM)
(Glintborg et al. 2012, Johnstone et al. 2012, Puurunen et al. 2011) . However, the
mechanisms leading to these increased metabolic risks, especially those regarding
lifelong morbidity and mortality linked to the syndrome are still controversial, as
the current evidence has come mostly from cross-sectional data sets with relatively
small numbers of subjects and heterogeneous populations (Dumesic et al. 2015).
Anti-Müllerian hormone (AMH) is involved in sexual differentiation
prenatally and later in ovarian follicle growth. It is produced by small ovarian
follicles and its serum levels reflect the function of the ovary. It is well known that
AMH levels are high in women with PCOS (Dewailly et al. 2010), but follow-up
studies on the capacity of AMH to predict the development of PCOS are still
lacking (La Marca et al. 2009, Piltonen et al. 2005).
Given the reproductive and metabolic risks linked to PCOS, it is important to
identify these women early in order to prevent morbidity and even mortality linked
to these risks throughout life. Improved knowledge concerning the prediction of
PCOS would help clinicians to design tailored treatment to prevent late health
disorders linked to the syndrome.
The main objective of the present study was to provide new information on the
relationships between obesity, testosterone levels and symptoms of
19
hyperandrogenemia in cases of PCOS both in adolescence and in early adulthood.
A second objective was to study the value of AMH in adolescence as a possible
PCOS marker, and its association with reproductive and metabolic health later in
early adulthood.
Lastly, we wanted to investigate the evolution of hormonal and metabolic
parameters with age in both control women and women with PCOS and to find cutoff values of hormonal parameters in regard to the risk of PCOS at a population
level.
Our hypothesis was that hyperandrogenemia and/or menstrual irregularities in
adolescence have a major impact on women’s reproductive and general health both
in adolescence and in later life. In addition, we hypothesized that hormonal and
metabolic changes with age occur differently in women with PCOS compared with
a control population.
20
2
Review of the literature
2.1
Androgen secretion in women
Approximately 25% of the testosterone in women is produced by the ovaries,
another 25% by the adrenal glands and the remaining 50% through peripheral
conversion of estrogens to androgens (Burger 2002, Longcope 1986). The major
androgens in women, listed in decreasing order according to serum concentrations,
are dehydroepiandrosterone sulfate (DHEAS), dehydroepiandrosterone (DHEA),
androstenedione (A4), testosterone (T) and dihydrotestosterone (DHT). Of these, T
and DHT are considered to induce most of the androgenic effects by binding to
nuclear androgen receptors, while A4, DHEA and DHEAS are pro-androgens
requiring conversion to T and DHT in order to have an androgenic effect (Burger
2002). Androgen secretion in women originates in response to their tropic
hormones regulating it, i.e. luteinizing hormone (LH) acting on the ovary and
adrenocorticotropic hormone (ACTH) on the adrenal gland (Figure 1). Virtually all
of the circulating DHEAS in women is of adrenal origin (Longcope 1986, Stanczyk
2006). Both in the ovaries and in the adrenals, cholesterol is the precursor of
pregnenolone, which is further converted to androgens through an enzymatic
cascade described in Figure 2.
The rate-limiting step in androgen production is the movement of cholesterol
inside the cell brought about by steroidogenic acute regulatory protein (StAR). The
conversion of cholesterol to pregnenolone is regulated by tropic hormones and
carried out by the side-chain cleavage enzyme (P450scc). As regards steroid
hormone synthesis, the rate-limiting step is regulation of P450c17 gene (17hydroxylase and 17,20 lyase activity) expression inducing DHEA and A4
production from pregnenolone and progesterone. Furthermore, the expression of
P450c17 depends on the concentrations of tropic hormones in the ovaries and
adrenal glands. Other important enzymes include 3β-hydroxysteroid
dehydrogenase, inducing DHEA conversion to A4, and 17β-hydroxysteroid
dehydrogenase catalyzing A4 conversion to T (Baptiste et al. 2010 and Figure 2).
The most potent endogenous circulating androgen is T and its biological
activity depends on the amount of its free fraction in the blood. Most of the
circulating T is bound to carrier proteins such as sex hormone-binding globulin
(SHBG) and (partly) albumin, leaving only about 1–2% circulating unbound
(Catteau-Jonard & Dewailly 2013). Therefore, changes occurring in SHBG levels
21
alter the severity of hyperandrogenism. As many factors (see 2.2.1) influence the
levels of SHBG, measurement of total testosterone alone may not be reliable in
estimating the degree of hyperandrogenism.
Fig. 1. The neuroendocrine axis regulating androgen biosynthesis in the ovaries and
adrenal glands.
22
Fig. 2. Female steroidogenetic cascade in the ovaries and adrenal glands. In the adrenal
gland steroidogenesis takes place in the inner parts of the adrenal cortex and in the
ovary the production of androgens takes place in the theca cells and estrogen
biosynthesis in the granulosa cells.
23
2.1.1 Ovarian androgen secretion
Gonadotropin-releasing hormone (GnRH) secreted from the hypothalamus causes
the pituitary to secrete LH and FSH, further controlling ovarian androgen and
estrogen biosynthesis. Ovarian steroidogenesis works in a two-cell model such that
LH stimulates the theca cells to produce androgens. Androgens then diffuse to the
FSH-controlled granulosa cells where they are aromatized to estrogens that again
have negative feedback on the hypothalamus, constituting a regulatory loop (Figure
1). Ovarian androgens are not part of a specific negative feedback system on
hypothalamic GnRH and pituitary LH secretion, as ovarian androgens are byproducts of estrogen biosynthesis. Therefore, the control of ovarian
hyperandrogenism in connection with normal estrogen synthesis is weak.
Serum testosterone levels are at their lowest in the early follicular phase of the
menstrual cycle. They rise to a peak at mid-cycle and decrease again during the
luteal phase towards the early follicular phase (Abraham et al. 1974, Bui et al.
2013).
According to the results of previous studies, a decrease in androgen levels in
women occurs during fertile life (Liang et al. 2011, Piltonen et al. 2003, Winters et
al. 2000). Thereafter, even though estradiol levels clearly decrease during
menopausal transition due to follicle depletion, little change is apparent in serum
androgen levels (Davison et al. 2005). After menopause androgen production
capacity in the ovaries diminishes progressively, but still persists for years
(Couzinet et al. 2001, Hughes et al. 1991) and there is evidence that T
concentrations in the ovarian vein are higher than those in the peripheral blood after
menopause. Thus, the ovaries appear primarily to be androgen-secreting organs
after the fertile period of life (Couzinet et al. 2001, Judd et al. 1974).
2.1.2 Adrenal androgen secretion
The three layers of the adrenal cortex each have specific enzymatic cascades
resulting in the biosynthesis of steroid hormones. The zona glomerulosa has the
capacity to produce mineralocorticoids, e.g. aldosterone. The inner layers, the zona
fasciculata and the zona reticularis, synthesize androgens (DHEA and A4). The
zona fasciculata also produces glucocorticoids such as cortisol, which is mainly
responsible for controlling ACTH levels via negative feedback (Figure 1) (Baptiste
et al. 2010). The adrenal medulla produces catecholamines (adrenaline and
24
noradrenaline), which are known as fight-or-flight hormones and are driven by the
sympathetic nervous system.
The most important adrenal androgens are DHEA and its sulfate conjugate
DHEAS. The ovaries produce approximately 20% of total circulating DHEA, but
DHEAS originates from the adrenal glands and peripheral tissues (small intestine
and liver) after conversion from DHEA (Goodarzi et al. 2014). DHEAS is
commonly used as a marker of adrenal androgen production (Goodarzi et al. 2014,
Lobo et al. 1981). The circulating (serum) concentration of DHEAS is about 1000fold compared with the concentration of DHEA due to its high production rate and
low metabolic clearance rate (Goodarzi et al. 2014). Depending on the phase of the
menstrual cycle, one third to a half of the amount of A4 is produced in the ovaries.
Due to the aforementioned decline in ovarian androgen production during
menopausal transition, the adrenal glands are the major source of circulating
androgens after the menopause (Couzinet et al. 2001). However, DHEAS levels
also decrease linearly with age (Goodarzi et al. 2014).
2.1.3 Peripheral tissues
Clinically the most important peripheral tissues converting androgens are the liver,
the lungs, adipose tissue and the skin (Ala-Fossi et al. 1998). These tissues present
both 3β-HSD and 17β-HSD enzyme activity in order to convert DHEA and
DHEAS into A4 and further to T (Payne & Hales 2004). Moreover, T is converted
to DHT by 5α-reductase, present in these peripheral tissues. The aromatase activity
of tissues such as the skin and adipose tissue enables the metabolism of androgens
to estrogens (Nelson & Bulun 2001).
2.2
Hyperandrogenism in women
Hyperandrogenism or androgen excess is a common endocrine condition affecting
approximately 5–10% of women of reproductive age (Azziz et al. 2004, Carmina
et al. 2006). There are multiple causes of hyperandrogenism in women and the most
common is PCOS, covering over 80% of women with androgen excess (Azziz et
al. 2004). The next most important causes of androgen excess are presented in
Table 1 (Azziz et al. 2004, Yildiz 2006). There are few well-characterized
hyperandrogenic syndromes (Cushing’s syndrome, androgen-secreting tumors and
non-classical adrenal enzymatic deficiencies) to be taken into account (Carmina et
al. 2006). Another possible cause of hyperandrogenemia is androgen intake.
25
Table 1. Causes of hyperandrogenic disorders in women. Modified after Carmina et al.
2005.
Hyperandrogenic disorders
Prevalence (%)
PCOS
72.1
Idiopathic hyperandrogenism
15.8
Idiopathic hirsutism
7.6
Non-classical congenital adrenal hyperplasia
4.3
Ovarian or adrenal androgen-secreting neoplasms
0.2
Clinically, hyperandrogenism in women presents as hirsutism, acne and androgenic
alopecia. However, it is noteworthy that not all women with hirsutism have
hyperandrogenism, and, vice versa, hyperandrogenic Asian patients with PCOS,
for example, exhibit less hirsutism and acne (Carmina et al. 1992).
Biochemically, hyperandrogenism is defined as elevated levels of T and
calculated indices of androgenicity, such as the free androgen index (FAI) and free
testosterone (FT) exceeding the upper limit of the normal range (see 2.2.1).
2.2.1 Evaluation of hyperandrogenism in women
Clinical evaluation
Hirsutism is the most reliable symptom used in clinical practice for the evaluation
of clinical hyperandrogenism, whereas acne and alopecia show poor correlation
with it. Androgens determine the amount and distribution of hair-growth, clinically
expressed as hirsutism, i.e. excess male-pattern terminal hair growth. Hirsutism is
a clinical diagnosis and is generally evaluated by using the Ferriman–Gallwey
(F&G) scale, which gives scores from 1 to 4 for the presence of hair growth in
eleven areas of the body (Ferriman & Gallway 1961, Table 2). An F&G score of >7
defines hirsutism.
26
Table 2. Definition of hair growth grading according to Ferriman and Gallwey.
Site
Grade
Definition
1. Upper lip
1
A few hairs at outer margin.
2
A small mustache at outer margin.
3
A moustache extending halfway from outer margin.
4
A moustache extending to mid-line
2. Chin
3. Chest
4. Upper back
5. Lower back
6. Upper abdomen
7. Lower abdomen
8. Arm
1
A few scattered hairs.
2
Scattered hairs with small concentrations.
3&4
Complete cover, light and heavy.
1
Circumareolar hairs.
2
With mid-line hair in addition.
3
Fusion of these areas, with three-quarter cover.
4
Complete cover
1
A few scattered hairs.
2
Rather more, still scattered.
3&4
Complete cover, light and heavy.
1
A sacral tuft of hair.
2
With some lateral extension.
3
Three-quarter cover.
4
Complete cover.
1
A few mid-line hairs
2
Rather more, still mid-line.
3&4
Half and full cover.
1
A few mid-line hairs.
2
A mid-line streak of hair.
3
A mid-line band of hair.
4
An inverted V-shaped growth
1
Sparse growth affecting not more than a quarter of the limb
surface.
2
More than this; cover still incomplete.
3&4
Complete cover, light and heavy.
9. Forearm
1, 2, 3 & 4 Complete cover of dorsal surface; 2 grades of light and 2 of
10. Thigh
1, 2, 3 & 4 As for arm
11. Leg
1, 2, 3 & 4 As for arm
heavy growth
At present, other scoring systems to diagnose hirsutism are also used, such as
modifications of the original F&G score, using five to nine of the body areas
originally assessed by Ferriman and Gallwey (Yildiz 2006, Figure 3). Ethnic
variations should be taken into account when assessing hirsutism, e.g. Asians have
less dense hair than Caucasians (Ewing & Rouse 1978). The subjectivity of self27
reported hirsutism may cause unreliability in reporting clinical hyperandrogenism,
even though self-reported hirsutism has been shown to identify most women with
typical PCOS findings (Kazemi et al. 2015, Taponen et al. 2003, Taponen et al.
2004).
Fig. 3. Modified Ferriman & Gallwey scoring sheet used for clinical evaluation of hair
growth. Scores of seven or more are defined as hirsutism. Modified from Hatch et al.
1981.
Acne and alopecia are diagnosed according to a carefully recorded patient history
and physical examination. Acne has a multifactorial etiology, but androgens play
an important role in the development of this disorder. The androgenic profile should
be examined in patients suffering from severe disease or additional symptoms of
hyperandrogenism (Yildiz 2006). Alopecia, i.e. loss of scalp hair, is caused by
increased androgen activity in the hair follicles. Most women suffering from it,
28
however, do not exhibit increased levels of circulating androgens and its
significance in the diagnosis of PCOS is low (Rebora 2004, Schmidt & Shinkai
2015).
Virilization is a relatively uncommon finding, extensively including clinical
features of hyperandrogenism such as hirsutism, acne, alopecia, clitoromegaly,
deepening of the voice, loss of female body shape and usually amenorrhea. In
women with PCOS mild virilization (mostly solely clitoromegaly) has been
described as having an incidence of 21%, and other studies on the prevalence of
this alteration are scarce (Goldzieher & Axelrod 1963). Women with virilization
usually present markedly increased levels of androgens (Marshburn & Carr 1995)
and androgen-secreting tumors should always be ruled out (Fraser & Kovacs 2004).
Biochemical evaluation
In order to complete the evaluation of hyperandrogenism the clinician needs to
understand the methods used to measure androgens in the clinical laboratory. The
methods in use vary as regards diagnostic accuracy, strengths and limitations.
Especially when measuring the relatively low levels of androgens in women, the
method must be relevant and highly reliable (Stanczyk 2006).
Direct immunoassay, the most commonly used measurement method for serum
concentrations of T, is generally considered to be inaccurate, particularly when
measuring the low levels found in women and children (Handelsman & Wartofsky
2013). Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is
available for use in T measurement, but initially it was too expensive for routine
use. This has changed during the last decade and LC-MS/MS has become the goldstandard method for T measurement. Many journals in the field nowadays
recommend this method in endocrine studies (Handelsman & Wartofsky 2013,
Janse et al. 2011). The strengths of LC-MS/MS are enhanced analytical specificity,
sensitivity and accuracy and it serves as a reference method for a number of
analytes (Shea et al. 2014). However, a comparative study of two assays showed
that in women with PCOS a select radioimmunoassay (RIA) method and LCMS/MS had equal precision in measuring T and that correlation with the clinical
parameters may be even stronger with RIA (Legro et al. 2010b).
DHEAS is present at high concentrations in female serum and direct
immunoassay is considered to be reliable. Androstenedione and DHEA should
preferably be measured by LC-MS/MS, unless the direct immunoassay methods
used have been extensively validated (Stanczyk 2006).
29
Sex hormone-binding globulin
Sex hormone-binding globulin is a glycoprotein encoded by a single gene in
chromosome 17 and it is produced in the liver by hepatocytes (Hammond &
Bocchinfuso 1996). SHBG has a unique capacity to bind sex hormones with high
affinity, which decreases their metabolic clearance and regulates their
bioavailability (Joseph 1994). Serum levels of SHBG are regulated mainly by sex
hormones, but thyroid hormones, cytokines and some dietary components have also
been shown to be involved in the regulation of SHBG production (Simo et al. 2015,
Thaler et al. 2015). Therefore, some disorders such as pituitary, thyroid or liver
diseases, as well as breast cancer, may influence the production of SHBG. Changes
in circulating levels of SHBG affect the amount of biologically active free
androgens, thus affecting the clinical presentation linked to the severity of hypo- or
hyperandrogenism. Hyperinsulinemia due to insulin resistance decreases the
synthesis of SHBG and its levels in plasma, whereas weight loss or use of insulin
sensitizers increase serum SHBG levels (Pasquali et al. 1995, Plymate et al. 1988).
However, recent studies in vitro and in vivo could not confirm the mechanism of
insulin-regulated SHBG production and further studies are required (Simo et al.
2015, Winters et al. 2014). Further, it has been shown that the strongest predictor
of SHBG levels is the amount of liver fat and that there is an inverse correlation
between the two parameters (Peter et al. 2010).
Free Androgen Index and Calculated Free Testosterone
To overcome the limitations of T measurement previously described, algorithms
have been developed to estimate the amount of free T. The free androgen index
(FAI) is calculated from T and SHBG concentrations (Table 3) and it is considered
more informative than T alone as regards evaluation of androgenicity (Cho et al.
2008, Handelsman & Wartofsky 2013, Pinola et al. 2015). However, the FAI has
been shown to be unreliable in men, but it seems to correlate well with the level of
free testosterone in women (Miller et al. 2004, Rosner et al. 2007). The results of
some studies suggest that calculated free testosterone (cFT, Table 3), which also
takes into account albumin concentrations, could be more informative than the FAI
(Rosner et al. 2007, Shea et al. 2014). Calculated FT has been shown to correlate
well with free testosterone concentrations measured by equilibrium dialysis, which
is considered to be the most sensitive and physiologically adequate form of analysis
of T (Vermeulen et al. 1999). In the equation for cFT the albumin concentration is
30
fixed within the normal physiological range (at 43 g/L) and it has shown good
accuracy, except in pregnancy, when albumin concentrations are typically lower.
In the latest Rotterdam consensus the use of cFT or FAI instead of serum total
T was recommended in order to detect hyperandrogenism in women with PCOS
(Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group 2004).
The reference ranges for FAI and cFT have to be verified and validated in each
laboratory in different populations as a result of the different methods used for T
and SHBG analyses (Salameh et al. 2014, Shea et al. 2014).
Table 3. Algorithms used for evaluation of adrogen indices.
Androgen indices
Equation
Free Androgen Index (FAI)
FAI = 100 × (total T / SHBG)
Calculated Free Testosterone
cFT = {-b + √(b2 + 4a[TT]} / 2a
(cFT)
1
a = kat + kt + (kat × kt)([SHBG] + [albumin] – [T])
b = 1 + kt[SHBG] + kat[albumin] + (kat + kt)[T]
1
Vermeulen et al. 1999
2.3
Anti-Müllerian hormone
Anti-Müllerian hormone (AMH), also known as Müllerian inhibiting substance, is
a dimeric glycoprotein and a member of the transforming growth factor-beta (TGFβ) superfamily (La Marca et al. 2009). The TGF-β family includes a large number
of structurally related proteins with multiple roles in developmental patterning,
tissue differentiation, and maintenance of homeostasis (Morikawa et al. 2016). In
embryogenesis AMH plays a central role in sexual differentiation by inducing
regression of the Müllerian ducts in male fetuses. In females, AMH is produced by
the granulosa cells of the human ovary after midgestation in the fetus (Rajpert-De
Meyts et al. 1999, Vigier et al. 1984). AMH is expressed in granulosa cells of
growing follicles up to antral stage, suggesting an important role in early ovarian
folliculogenesis (Stubbs et al. 2005, Weenen et al. 2004). AMH is able to inhibit
the initiation of primordial follicle growth (Durlinger et al. 2002) and it may also
decrease the sensitivity of antral follicles to follicle-stimulating hormone (FSH)
(Durlinger et al. 1999, Durlinger et al. 2002, Gruijters et al. 2003).
In humans, the physiological relationship between circulating levels of
androgens and AMH remains uncertain, and the exact function of AMH in follicular
recruitment and its long-term effects are not well understood. Serum AMH levels
31
and the ovarian antral follicle count (AFC) correlate to each other both in healthy
subjects and in women with PCOS (Pigny et al. 2003, Pigny et al. 2006, Weenen
et al. 2004). In PCOS, serum concentrations of AMH correlate positively to those
of serum testosterone and negatively to age (Piltonen et al. 2005). Furthermore,
women with elevated serum levels of AMH and T have longer menstrual cycles
compared with those with lower levels (Kristensen et al. 2012). AMH may also
have an important role in regulation of the GnRH-dependent pulsatility of LH.
AMH receptors have been discovered in GnRH neurons of mice and humans, and
in murine studies AMH has been shown to activate GnRH neuron firing (Cimino
et al. 2016). This is of importance, as it might clarify the extragonadal effects of
AMH and add new information on the pathophysiology of PCOS in the future.
As AMH levels strongly correlate to both biochemical hyperandrogenism (HA)
and antral follicle count, it has been suggested that it could be used as a surrogate
tool in the diagnosis of PCOS instead of polycystic ovary morphology (PCOM, see
2.4.4) (Casadei et al. 2013, Chen et al. 2008, Eilertsen et al. 2012, Fanchin et al.
2003, Nardo et al. 2009). A serum AMH concentration over 35 pmol/L has been
suggested to be more reliable than antral follicle count (Dewailly et al. 2011).
However, up to now, AMH is not listed in the validated criteria for PCOS diagnosis
(Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group 2004).
There are few studies on the possible value of the AMH serum level as a diagnostic
tool for the diagnosis of PCOS in adolescence or as a possible predictor of PCOS
in later life, and these studies have involved relatively small numbers of subjects
and have produced controversial results (Pawelczak et al. 2012, Sopher et al. 2014).
AMH has also been shown to be associated with IR and to reflect disturbances
of gonadotropin release in women with PCOS. However, results regarding the
relationship between metabolic risk factors and AMH remain controversial (de Kat
et al. 2015) and further studies are needed to evaluate the importance of AMH as a
cardiovascular risk factor (Skalba et al. 2011). The process of ovarian aging
(expressed in terms of AMH decline) and the emergence of cardiovascular risk
factors seem to occur simultaneously during perimenopausal transition, and further
studies are needed to clarify the connective mechanism (de Kat et al. 2015).
32
2.4
Polycystic ovary syndrome
2.4.1 Definition and prevalence
The first international diagnostic criteria (Table 4) for PCOS were developed in
1990 by a group of scientists and clinicians attending a conference sponsored by
the National Institutes of Health (NIH). PCOS was defined as hyperandrogenism
and/or hyperandrogenemia together with oligo-anovulation after exclusion of other
endocrinopathies (Zawadski & Dunaif 1992). In 2003, PCOM was added as one of
the criteria and it was required that at least two criteria should be met for PCOS
diagnosis (Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop
group 2004).
Thereafter, in 2012 at an NIH workshop it was recommended that clinicians
use the Rotterdam criteria (NIH guidelines of PCOS 2012). In addition, the
Androgen Excess and PCOS (AE-PCOS) Society emphasized the importance of
HA in PCOS, in keeping with the NIH criteria from 1990 (Azziz et al. 2006).
PCOS therefore represents a multifaceted syndrome with different phenotypes
(Table 4). The prevalence of PCOS varies according to the criteria used, being
twofold according to the Rotterdam criteria (17.8%) compared with the NIH
criteria (8.7%) and with a prevalence of 12% according to the AE-PCOS society
criteria (March et al. 2010). However, the prevalence also varies depending on
ethnicity and the population studied. The generally accepted prevalence is between
6 and 18% (Carmina et al. 2006, Franks 1995, March et al. 2010).
Table 4. Different diagnostic criteria used for defining PCOS.
Criteria
Hyperandrogenism
Oligo-anovulation
PCOM
Rotterdam A* / AE-PCOS
+
+
+
Rotterdam B* / AE-PCOS
+
+
-
Rotterdam C* / AE-PCOS
+
-
+
Rotterdam D*
-
+
+
NIH
+
+
-
*Different phenotypes of PCOS according to the Rotterdam criteria.
33
2.4.2 Hyperandrogenism in PCOS
It is generally accepted that hyperandrogenism is a central feature in the syndrome.
From 58 to 82% of hyperandrogenic women have PCOS, depending on the
diagnostic criteria used (Azziz et al. 2004, Carmina et al. 2006).
In PCOS, the ovaries cover up to 60% of the total production of androgens.
After suppressing the ovarian production with a gonadotropin-releasing hormone
(GnRH) agonist, women with PCOS also exhibit adrenal overproduction of
androgens, and vice versa (Cedars et al. 1992). Combined with low levels of SHBG
linked to increased insulin resistance and accumulation of liver fat in PCOS, the
hypersecretion of androgens leads to excess levels of biologically active forms of
androgens.
Women with PCOS show an altered hypothalamic–pituitary axis, leading to
increased LH secretion as regards pulse frequency and amplitude, and causing
ovarian theca cells to overproduce androgens (Morales et al. 1996). Women with
PCOS also present a higher number of androgen-producing theca cells, with
increased expression of LH receptors (Gilling-Smith et al. 1994, Jakimiuk et al.
2001). In addition, androgenic hyper-responsiveness in PCOS seems to be linked
to insulin action, which possibly mimics the tropic action of LH on theca cells (Wu
et al. 2014). Further, improvement of IR in cases of PCOS has been shown to
decrease the level of HA (Baillargeon et al. 2004).
Adrenal hyperandrogenism is a frequent issue in PCOS. Additionally, several
extra-adrenal mechanisms have also been identified; of these, hyperresponsiveness
to ACTH stimulation and long-term effects of high LH levels seem to play an
important role (Cinar et al. 2012, Pabon et al. 1996, Tock et al. 2014). In women
with normal ACTH levels, increased LH secretion, as observed in PCOS, may
explain higher DHEAS concentrations, as LH/hCG receptors have been found in
the adrenals, resulting in increased steroidogenesis of the adrenal cortex (Pabon et
al. 1996). However, opposite results have also been published (Piltonen et al. 2002).
In a recent review it was also suggested that hyperinsulinemia and increased
peripheral metabolism of cortisol in women with PCOS exacerbate ACTH
synthesis in order to maintain sufficiently high cortisol levels, with the by-product
of increased adrenal androgen synthesis (Baskind and Balen 2016). However,
further studies on the relationships between adrenal function, PCOS and obesity
are needed.
Hyperandrogenism eases with time in women and this decreasing trend with
age has been observed in both healthy women and PCOS patients (Bili et al. 2001,
34
Hudecova et al. 2009, Hudecova et al. 2011, Piltonen et al. 2004, Purifoy et al.
1980, Ruutiainen et al. 1988). However, despite the steady decrease of androgens
with age, hyperandrogenism seems to persist or even worsen after menopause in
women with PCOS (Burger et al. 2000, Burger et al. 2007, Puurunen et al. 2009,
Puurunen et al. 2011, Schmidt et al. 2011, Winters et al. 2000). This phenomenon
may be a result of higher gonadotropin levels after the menopause, a decrease in
SHBG levels or a co-gonadotropic action of insulin. Overall, the mechanisms of
hyperandrogenism after menopause appear to be multiple and the correlation to
morbidities and mortality remains unclear (Markopoulos et al. 2015).
2.4.3 Menstrual irregularities in PCOS
The majority of the PCOS classifications include oligo-amenorrhea (OA) as a
component (Table 4). This is defined as a menstrual cycle longer than 35 days more
than twice a year, or chronic anovulation (Azziz et al. 2009, Broekmans & Fauser
2006). Among women with PCOS, OA is a very common feature and its prevalence
varies from 75 to 90% depending on the age of the women and the criteria used to
define menstrual irregularities (Adams et al. 1986, Burgers et al. 2010, Hull 1987).
Menstrual irregularities in adolescence are a frequent finding and are present
in 20–30% of girls. Importantly, primary amenorrhea (genetic abnormalities,
malformations)
and
secondary
amenorrhea
(thyroid
dysfunction,
hyperprolactinemia and eating disorders) need to be excluded. It is important to
understand that maturation of the hypothalamic–pituitary–gonadal axis has been
evaluated to take up to five years (Apter & Vihko 1983). However, previous studies
have shown that two thirds of girls suffering from menstrual irregularities two years
after menarche still present menstrual disorders 10 years later and tend to have
PCOM and hyperandrogenism (elevated levels of A4 and T) in early adulthood (van
Hooff et al. 2000, Wiksten-Almstromer et al. 2008). However, menstrual
irregularities seem to ease with age in women with PCOS (see p. 45).
As for metabolic features, previous studies have not shown any association
between OA and insulin resistance in adolescence, but girls of identical age and
fulfilling PCOS diagnosis criteria exhibited an approximately 50% reduction in
insulin sensitivity (Lewy et al. 2001, van Hooff et al. 2000). In adulthood, however,
the menstrual pattern has been proposed as a surrogate marker of the metabolic
profile in women with PCOS, as amenorrhea seems to be associated with more
pronounced IR and hyperandrogenemia (Panidis et al. 2013). All in all, the
35
relationship between menstrual irregularities, insulin resistance, metabolic
disorders, hyperandrogenemia and later risk of PCOS remains to be clarified.
2.4.4 PCO morphology at ultrasonography in PCOS
Polycystic ovary morphology was originally defined as the presence of ten or more
follicles with a diameter of two to eight millimeters and/or increased ovarian
volume (over 10 milliliters) in at least one ovary as observed in transabdominal
ultrasonography (Adams et al. 1986). The Rotterdam consensus defined PCOM as
the presence of 12 or more follicles with a diameter of two to nine millimeters
and/or increased ovarian volume (over 10 milliliters) in at least one ovary
(Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group 2004).
With improving imaging systems it has been debated whether this criterion should
be changed. In 2014 a panel of experts in the field recommended that PCOM should
be defined as 25 or more follicles in one ovary (Dewailly et al. 2014). However, so
far, the Rotterdam criteria are still in use in clinical practice.
It has to be noted that PCOM is a common finding among young healthy
women and that its prevalence has been estimated to be three- to fourfold when
compared with the prevalence of PCOS (Johnstone et al. 2010, Polson et al. 1988).
PCOM is also frequently found in adolescents shortly after menarche. It has been
suggested that in young women with a gynecological age over two years, the most
reliable indicator of PCOM may be ovarian enlargement (> 10 cm3, Carmina et al.
2010).
2.4.5 Other clinical features of PCOS
Obesity
Obesity is an increasing problem in the world. It is common in women with PCOS
but is probably not a cause of the syndrome, as PCOS is also seen among lean
women. Obesity seems to exacerbate many aspects of the syndrome (DiamantiKandarakis & Panidis 2007, Hoeger & Oberfield 2012, Legro 2012). It has been
proposed that lean women with PCOS have the most severe defect in ovarian
steroidogenesis and thus exogenous factors may not further provoke development
of PCOS, whereas among women with mild defects the contribution of obesity,
36
abdominal obesity or insulin resistance will lead to development of the full-blown
syndrome (Homburg et al. 2009, Escobar-Morreale & San Millan 2007).
The prevalence of obesity among women with PCOS varies greatly between
different populations, from about 80% in the United States to approximately 50%
in other countries (Ehrmann et al. 1999, Legro et al. 1999, Legro 2012).
Even though obesity is not a diagnostic criterion as regards PCOS, it exacerbates
many aspects of the syndrome, especially risk factors linked to cardiovascular
diseases, e.g. alterations in glucose and insulin metabolism, and dyslipidemia
(Moran et al. 2010). The physiopathology linking obesity and PCOS is complex
and one possible mechanism may be interaction between insulin resistance and the
co-gonadotropic effects of insulin and increasing circulating androgen levels
(Harrison et al. 2011, Moran et al. 2003). However, lifestyle modifications and
weight reduction are essential parts of the treatment of the syndrome, as they
improve the menstrual cycle and reduce IR and HA, independently of the type of
exercise (Harrison et al. 2011). Among morbidly obese women, however, the
prevalence of PCOS is only 13%, suggesting that there are other factors in addition
to obesity explaining the development of the syndrome (Gosman et al. 2010, Legro
2012).
Insulin resistance (IR)
When tissues normally sensitive to glucose-stimulating effects (e.g. muscle, liver
and adipose tissue) became resistant to these signals the state is called insulin
resistance (Smith 1994). Though IR and related compensatory hyperinsulinemia
are common findings in patients with PCOS, these are not included in the
diagnostic criteria. The relationship between HA, IR and obesity (Figure 4) is
complex and despite extensive studies is still not totally clarified (Dumesic et al.
2015). Insulin resistance leads to impaired glucose tolerance (IGT), which is
present in up to 30% of women with PCOS at reproductive age (Legro et al. 1999).
Up to 10% of women with PCOS will develop T2DM before the age of 40 years,
and obesity substantially increases (doubles) this risk (Legro et al. 1999, Moran et
al. 2010).
Altered insulin metabolism is present both in lean and obese women with PCOS,
but it is most marked when obesity, especially the abdominal type, is present
(Diamanti-Kandarakis et al. 1995, Dunaif et al. 1992, Legro et al. 1999, MorinPapunen et al. 2000). In the past 20 years there has been growing evidence
supporting the hypothesis that altered insulin action and insulin signaling pathways
37
play an important role in the pathogenesis of PCOS (Baptiste et al. 2010).
Furthermore, the use of insulin sensitizers, such as metformin, has been associated
with improvement of menstrual cyclicity and hyperandrogenemia (Morin-Papunen
et al. 1998). Additionally, according to some (Morin-Papunen et al. 2012), but not
all studies (Legro R et al. 2007), metformin may be useful as an adjuvant treatment
to achieve ovulation and live birth in obese women with PCOS and anovulatory
infertility. However, the nature and mechanism of the disorders of insulin action in
PCOS are still not fully clarified and further research is needed.
Chronic hyperinsulinemia seems to play an important role in the modulation
of CVD risk factors and it also exacerbates insulin resistance and directly
contributes to β-cell failure and development of diabetes (Randeva et al. 2012).
Importantly, insulin resistance is linked to CVD risk factors and general health
problems, such as metabolic syndrome (MetS) (Ginsberg 2000).
The gold-standard method for evaluating insulin action and secretion is the
euglycemic hyperinsulinemic clamp (DeFronzo et al. 1979). Another highly
specific and sensitive method is the intravenous glucose tolerance test (Bergman et
al. 1989). However, these methods are time-consuming and expensive and
therefore oral glucose-stimulated assessments (e.g. homeostasis model assessment
for insulin resistance [HOMA-IR, calculated as: fasting glucose × fasting insulin /
22.5] and the Matsuda index [calculated as: 10000 / (fasting glucose × fasting
insulin × mean OGTT glucose × mean OGTT insulin)0.5]) for evaluating insulin
resistance are generally used clinically (Ader et al. 2014, Diamanti-Kandarakis et
al. 2004, Hucking et al. 2008).
38
Fig. 4. Interactions
between
abdominal
obesity,
insulin
resistance
and
hyperandrogenemia in the pathogenesis and progression of PCOS. Modified after
Rojas et al. 2014.
2.4.6 Metabolic risks in PCOS
Various metabolic risks have been reported in patients with PCOS. The most
common metabolic alteration is dyslipidemia (Hoffman & Ehrmann 2008). Levels
of (cardioprotective) high-density lipoprotein (HDL) are decreased whereas those
of low-density lipoprotein (LDL) and triglycerides (TGs) are elevated in women
with PCOS (Hoffman & Ehrmann 2008). Both lean and obese women with PCOS
exhibit an unbeneficial lipid profile compared with healthy, age- and BMI-matched
control women (Glueck et al. 2009, Yildirim et al. 2003). However, the ethnic
background (lifestyle and nutrition) as well as genetic differences may have an
effect on the characteristics of dyslipidemia, as, for example, it seems that
hypertriglyceridemia is more prevalent in Caucasian populations (Essah et al. 2008,
Hillman et al. 2014, Wild et al. 2010). Together with other metabolic disturbances,
dyslipidemia seems to be already present in adolescent subjects with PCOS (Kent
& Legro 2002). Moreover, HA seems to be a key factor associated with the most
adverse lipid profile in women with PCOS at a young age (Fruzzetti et al. 2009).
Circulating levels of the most commonly used marker of chronic inflammation,
C-reactive protein (CRP), are elevated in women with PCOS (Escobar-Morreale et
al. 2011). However, a previous meta-analysis did not reveal any differences
between women with PCOS and healthy women as regards other inflammatory
39
markers, such as interleukin-6 and tumor necrosis factor-alpha (Escobar-Morreale
et al. 2011). According to a recent review, studies on other adipokines (adiponectin,
resistin and visfatin) in PCOS are also inconclusive and more studies are needed
(Spitzer et al. 2015). Inflammation markers have been shown to be related to an
increased risk of cardiovascular events independently of other risk factors (Danesh
et al. 2004). Whether or not alterations in the levels of inflammation markers play
a causal role in the syndrome or are a consequence of other metabolic abnormalities
needs to be further clarified.
Women with PCOS have a higher prevalence and elevated risk of hypertension,
early-onset atherosclerosis and cardiac dysfunction (Elting et al. 2001, Prelevic et
al. 1995). The risk of developing arterial hypertension has been shown to be
increased 2.5-fold at menopausal age compared with a healthy age-matched
population, but it seems to be at least partly related to the obesity associated with
the syndrome (Elting et al. 2001). In one study, young subjects with PCOS (aged
22 yr) already exhibited an increase in carotid intima-media thickness, a surrogate
marker of atherosclerosis, suggesting early development of atherosclerosis in
PCOS (Orio et al. 2004).
Despite the various CVD risks linked to PCOS, definitive data on CVD-related
mortality in women with PCOS is lacking. In a follow-up study lasting 31 years,
no significantly increased risk of CVD-related causes of death were found, even
though nonfatal cerebrovascular diseases were more prevalent in women with
PCOS (Wild et al. 2000). However, in a large prospective cohort, women reporting
a history of menstrual irregularities had increased risks of both nonfatal and fatal
CVDs (Solomon et al. 2002). Long-term prospective menopausal data have shown
a higher incidence of hypertension and elevated triglyceride levels but no increased
rates of cardiovascular events (Schmidt et al. 2011). Further, PCOS has been shown
to be related to increased thrombin generation, which is considered to be a CVD
risk marker, and the use of oral contraceptives in the treatment of the syndrome has
even worsened this alteration (Glintborg et al. 2015a, Glintborg et al. 2015b). The
2012 Amsterdam Consensus Workshop Group opinion concerning women’s health
in PCOS was that longitudinal studies are needed to clarify the associations
between CVD risk markers and vascular events (Amsterdam ESHRE/ASRMSponsored 3rd PCOS Consensus Workshop Group 2012).
The Rotterdam Consensus Group (Rotterdam 2004) has defined criteria for the
diagnosis of MetS in PCOS populations (Table 5). The prevalence of MetS in PCOS
patients varies from 33–47% in the U.S. to 8–25% in other countries, which
represents a two- to threefold increase compared with age-matched controls (Wild
40
et al. 2010). Previous observations link hyperandrogenism to metabolic
disturbances and a higher prevalence of MetS (Sung et al. 2014) and according to
the Amsterdam consensus on the metabolic risks linked to PCOS, there is extensive
evidence showing that MetS is most prevalent in the hyperandrogenic and
anovulatory phenotype of PCOS. Thus, the consensus group recommends that
patients in this clinical subset should be treated carefully, keeping in mind the
possible long-term health risks (Amsterdam ESHRE/ASRM-Sponsored 3rd PCOS
Consensus Workshop Group 2012). On the other hand, the evidence in a large
review supported the postulate that obesity, especially abdominal, is associated
with a more adverse metabolic status (Moran & Teede 2009). Further, it is
noteworthy that as early as in adolescence, women with PCOS and normal BMI
exhibit a higher prevalence of MetS compared with control women, highlighting
the importance of identifying young women at risk (Aydin et al. 2015). The
Androgen Excess and Polycystic Ovary Syndrome (AE-PCOS) Society
recommends lifestyle management for primary CVD prevention, targeting
treatment of obesity, insulin resistance, elevated blood pressure and dyslipidemia
(Wild et al. 2010).
Table 5. Criteria for metabolic syndrome in women with PCOS. Three out of five qualify
for a diagnosis of metabolic syndrome. (Modified from the Rotterdam consensus
statement for diagnosing PCOS [Rotterdam 2004]).
Risk factor
Cut-off
Waist circumference
> 88 centimeters
Triglycerides
≥ 1.69 mmol/L
High-density lipoprotein
< 1.29 mmol/L
Blood pressure
≥ 130 / ≥ 85 mmHg
Fasting and 2-hour glucose from oral glucose
6.11–6.99mmol/L and/or
tolerance test
2-h glucose 7.77–11.05 mmol/L
Overall, the roles and interactions of obesity, especially abdominal obesity,
hyperandrogenism and the syndrome per se on the development of cardiovascular
risks, events and mortality are still under debate.
2.4.7 Non-alcoholic fatty liver disease (NAFLD) and obstructive sleep
apnea (OSA) in PCOS
Women with PCOS exhibit a high prevalence of non-alcoholic fatty liver disease
(NAFLD) (Cerda et al. 2007). In NAFLD, insulin resistance has a key role and is
41
an independent contributor to progression of this condition (Mendez-Sanchez et al.
2007). NAFLD may develop at a young age, and liver function should be screened
at an early age in women with PCOS, particularly in obese subjects and in women
suffering from MetS (Randeva et al. 2012).
The prevalence of obstructive sleep apnea (OSA) has also been observed to be
higher in women with PCOS compared with women without it (Vgontzas et al.
2001). Many physiopathological mechanisms have been proposed to explain the
elevated risk of OSA in PCOS but no single validated mechanism has been
identified (Randeva et al. 2012). Clinicians managing patients with PCOS should
be aware of the high prevalence of OSA among these women and systematically
screen them for the condition.
2.5
PCOS from adolescence to postmenopause
Polycystic ovary syndrome is a lifelong syndrome. Its characteristics are well
known in the fertile, menopausal and even postmenopausal periods, but the
developmental origins and the characteristics of the syndrome in adolescence are
still to be clarified. The predominant features of the syndrome during life are
presented in Figure 5.
Fig. 5. Phenotypic features of women with PCOS during their lives. Modified from Welt
and Carmina 2013.
42
2.5.1 Prepubertal origins of PCOS
Increasing evidence shows that the intrauterine environment has an effect on
genetic programming later in life. The first evidence of this was introduced by
Barker et al. in 1989 and it is now known as the Barker hypothesis (Barker et al.
1989, Hanson & Gluckman 2008). Exposing a female fetus to high T levels
prenatally causes dose-related masculine-type behavioral traits in young girls
(Hines et al. 2002). Prenatally androgenized female Rhesus monkeys have been
shown to express phenotypic characteristics similar to those typical of PCOS
(Abbott et al. 1998, Abbott et al. 2002). In early gestation, androgenized female
fetuses of Rhesus monkeys show hypersecretion of LH, indicating early
programming of hypothalamic action. Later, they develop symptoms of PCOS
(hyperandrogenism, menstrual irregularities, PCOM) as well as metabolic features
of the syndrome in early life (Abbott et al. 2005). Many different hypotheses have
been proposed to explain the epigenetic phenomenon of an early developmental
origin of PCOS, but no validated data has yet been presented. In summary, studies
show that exposure to excess androgens in utero encourage excess androgen
production later in life. Some retrospective studies have also shown that girls born
small for gestational age are at a higher risk of later development of early pubarche,
early menarche and PCOS (Ibanez et al. 2007), but controversial results have also
been published (Laitinen et al. 2003, Legro et al. 2010a).
Early pubarche and adrenarche have been linked to later development of PCOS
(Ibanez et al. 2000). Such girls may be at risk of PCOS-like symptoms and MetS,
especially if they are obese (Ibanez et al. 1998, Zimmet et al. 2007). Prepubertal
daughters of women with PCOS exhibit elevated AMH levels, which are
considered to be a surrogate marker of the amount of antral follicles in the ovaries
and, further, a correlative marker of hyperandrogenism (Dewailly et al. 2010, SirPetermann et al. 2006, Visser & Themmen 2005).
2.5.2 PCOS in adolescence
During adolescence and puberty the symptoms of PCOS overlap with normal
physiological changes (Carmina et al. 2010). Menstrual irregularities are frequent
in adolescents and the hypothalamus–pituitary–ovary (HPO) axis may need 2–5
years after menarche to reach full maturation (Apter 1980, Welt & Carmina 2013).
In cases of PCOS ovarian volume and follicle number overlap with the volume and
number observed in healthy subjects in adolescence and early adulthood
43
(Mortensen et al. 2009). The diagnosis of hyperandrogenism is difficult in
adolescence, as acne is a common problem in young adults and hirsutism may take
several years before development to its full expression. For this reason, the most
useful diagnostic tool is to measure (elevated) androgen levels to define
hyperandrogenism in adolescence (Blank et al. 2008). Reaching a diagnosis of
PCOS in adolescence requires all three cardinal symptoms to be present. However,
if ultrasonography is not available the diagnosis can be made in the presence of
oligo-anovulation and biochemical hyperandrogenemia (Legro et al. 2013).
Age at menarche in girls with PCOS ranges from nine years upwards
(Dramusic et al. 1997, Nduwayo et al. 1992, Rachmiel et al. 2008, Welt & Carmina
2013). Earlier menarche is strongly related to obesity and later menarche to higher
androgen levels (Dramusic et al. 1997, Rosenfield et al. 2009). In a retrospective
analysis, women with PCOS were more likely to report earlier or later menarche
than age-matched controls (Carroll et al. 2012).
2.5.3 PCOS at reproductive age
Polycystic ovary syndrome generally remains stable in the early reproductive years
between the ages of 18 to 35 (Davison et al. 2005). During this period women with
PCOS may exhibit all the cardinal symptoms of the syndrome. Studies have shown
that the hyperandrogenic phenotypes in early adult life exhibit the most severe
disease, with not only anovulatory infertility and a higher risk of pregnancy
complications such as gestational diabetes mellitus, hypertensive disorders of
pregnancy and risk of pre-term birth, but also a worse metabolic profile (Barber et
al. 2007, Boomsma et al. 2006, Franks 1995, Han et al. 2011). However, some data
have shown that in overweight women with PCOS, serum concentrations of T are
not independently associated with a higher prevalence of MetS and that the
interaction between obesity, HA and PCOS is complex and still needs clarification
(Moran et al. 2013).
During the reproductive years serum androgen levels decrease steadily in
women with PCOS, with a decline in both ovarian and adrenal androgen production
capacity (Davison et al. 2005). At later reproductive ages women with PCOS
present lower F&G scores and lower T, A4 and DHEAS levels compared with
younger women, but these parameters seem to remain higher than in healthy agematched subjects (Bili et al. 2001, Carmina et al. 2012, Puurunen et al. 2009,
Winters et al. 2000).
44
Ovulatory function appears to improve with age in women with PCOS. Studies
have shown an increase of up to 30% in menstrual frequency with increasing age
and it has been suggested that a relative increase in FSH with age might induce
follicle development in later reproductive life in these women (Alsamarai et al.
2009, Carmina et al. 2012, Elting et al. 2000). The recovery of regular menstruation
with aging in cases of PCOS could be predicted by measurement of serum AMH in
a prospective dataset (5-year follow-up), showing that a serum AMH level lower
than 0.56 pmol/L was correlated with recovery of ovulatory function in
premenopausal women (Carmina et al. 2012).
Follicle count and ovarian volume exhibit a steady decline with age in all
women, thus decreasing the prevalence of PCOM in women with PCOS (Alsamarai
et al. 2009). Importantly, however, the number of follicles remains higher in women
with PCOS compared with healthy controls at all ages (Aiyappan et al. 2016).
2.5.4 PCOS in premenopause and menopause
There are currently no diagnostic criteria for PCOS at menopausal age, but
clinicians define it by a history of prior oligo-amenorrhea and hyperandrogenism
(Azziz et al. 2006). Previous studies have shown that enhanced androgen secretion
remains in perimenopausal and postmenopausal women with PCOS (Markopoulos
et al. 2011, Puurunen et al. 2009, Puurunen et al. 2011). The enhanced secretion
decreases with age and stabilizes, meeting the levels in healthy women at around
the age of 70 years (Schmidt et al. 2011).
Abdominal obesity, IR, chronic inflammation and dyslipidemia have been
shown to worsen throughout the menopausal transition in healthy women
(Matthews et al. 2009). Longitudinal data have shown that a history of androgen
excess and menstrual irregularity might not be associated with the increased
incidence of MetS during menopausal transition in healthy women (Polotsky et al.
2014). There is strong evidence that women with PCOS have more components of
MetS starting at an earlier age, therefore exposing these women for a longer period
of life to adverse CVD risk factors (Polotsky et al. 2014, Welt & Carmina 2013). It
is possible that the CVD risk factors become milder with age in women with PCOS.
Further studies are needed to clarify whether the subgroup maintaining high
androgen levels after menopause remains at a higher risk of cardiovascular
morbidity and mortality (Welt & Carmina 2013).
45
46
3
Purpose of the present study
Polycystic ovary syndrome represents a major health issue among fertile-aged
women, with adverse effects on women´s health integrity, affecting not only
reproductive function, but also the emergence of the most frequent causes of
morbidity and mortality in postmenopausal years, T2DM and cardiovascular
diseases (Daan et al. 2014, Franks 1995, Johnstone et al. 2012, Morin-Papunen et
al. 2000). Obesity and hyperandrogenism are part of the syndrome and
confounding factors regarding the health impact of PCOS in the long run. The
current epidemic of obesity suggests that the prevalence of PCOS and its adverse
health effects may rise further in the future.
In adolescence, irregular menstrual cycles, acne and hirsutism are the principal
symptoms of PCOS. Data on causality between clinical manifestations of
hyperandrogenism and later metabolic risks is emerging. Studies on the metabolic
significance of biochemical hyperandrogenism have shown controversial results,
with only few studies involving the most accurate methods for testosterone
determination.
Age has a significant effect on the syndrome as regards both hormonal and
metabolic profiles. Ethnicity also has an impact on the clinical presentation of the
syndrome. Up to now, there have been no studies describing the endocrine or
metabolic profiles of women with PCOS in large Nordic populations at different
ages.
Anti-Müllerian hormone is produced in the growing follicles of the human
ovary and it is under intensive study, as women with PCOS have higher
concentrations of serum AMH compared with healthy women. There are only a few
studies on the predictive and diagnostic value of serum levels of AMH as regards
PCOS and data on its relationship with metabolic risks in PCOS is still
controversial.
There is a need to clarify the roles and interactions of obesity, weight gain,
hyperandrogenism and the syndrome per se on the development of long-term health
disorders, such as MetS, and on the risks of T2DM and CVD in PCOS. There is
also a need to define the health impact of PCOS in Nordic populations, as there are
important differences between ethnicities as regards clinical presentation and
metabolic risks linked to the syndrome. Given the aforementioned metabolic risk
factors associated with the syndrome, predicting PCOS early in life using simple,
inexpensive and reliable tools is essential to decrease and prevent late morbidity
47
and mortality linked to it. Lastly, the effect of menopause on the hormonal and
metabolic hallmarks of the syndrome is still under debate.
The aims of the study were:
1.
2.
3.
4.
5.
6.
48
To evaluate the correlations between serum concentrations of testosterone and
the free androgen index versus menstrual irregularities and metabolic
parameters in PCOS in adolescence (at the age of 16 years).
To compare serum AMH concentrations in girls with menstrual irregularities
and in controls at the age of 16 years and to evaluate the predictive value of
AMH as regards menstrual irregularities and symptoms/diagnosis of PCOS
over a follow-up period of 10 years in a population-based study.
To evaluate the association between serum levels of AMH and metabolic
parameters in adolescence.
To investigate serum androgen levels at different ages both in women with
PCOS and control women.
To determine cut-off values for serum levels of androgens to distinguish
women with PCOS and healthy women.
At different ages, to investigate metabolic parameters both in women with
PCOS and control women, to compare these parameters between the two
populations and to clarify the respective roles of hyperandrogenism and obesity,
especially abdominal obesity, as metabolic risk factors in PCOS.
4
Subjects and methods
4.1
Study populations and respective methods
A summary of the subjects, main study parameters and main results of the original
studies is shown in Table 7.
4.1.1 Northern Finland Birth Cohort 1986 (Studies I and II)
The prospective Northern Finland Birth Cohort 1986 (NFBC-1986), which has
been followed up since the fetal period, consists of 9362 mothers and their 9479
births (9432 children born alive), who had an expected date of birth between July
1, 1985 and June 30, 1986, from the two northernmost provinces of Finland. In
total, 4567 girls were born. The earliest data consists of pregnancy and antenatal
records and clinical data of the newborns and this has been supplemented by way
of postal questionnaires and clinical examination at the ages of seven and eight
years.
Questionnaire and investigations at the age of 16 years
In 2001–2002, when the children were 15–16 years old (mean age 15.5 years,
standard deviation [SD] 0.37 years), the adolescents and their parents each received
a large postal questionnaire (80 and 76% response, respectively). The questionnaire
included questions about puberty, smoking and use of alcohol, and one question
about the regularity and length of the menstrual cycle: “Is your menstrual cycle (the
interval from the beginning of one menstrual period to the beginning of the next
period) often (more than twice a year) longer than 35 days?” The girls who
answered “yes” to this question were considered to be suffering from menstrual
irregularities and were classified as “symptomatic”. The girls who answered “no”
were defined as “non-symptomatic”. In addition, 3373 girls (74%) underwent
clinical examination (including weight, height and waist-hip measurements) and
gave fasting blood samples.
After excluding twins and triplets, pregnant girls (n = 20), oral contraceptive
users and users of other forms of hormonal contraception and treatment (n = 377),
and subjects with incomplete data (n = 824), 2448 singleton females remained
eligible at the age of 16 years (Figure 6). This population was used in Study I.
49
Fig. 6. Flowchart of the study population in NFBC-1986 in Study I.
Questionnaire at the age of 26 years
At the age of 26 years, the young women (n = 4567) were sent another
questionnaire with questions on socio-demographic and other (health) background
factors mainly concerning reproduction, menstruation (including the same question
on menstrual disorders as at the age of 16 years) and infertility. After combining
data from the 16- and 26-year follow-up questionnaires, the final study sample
included 2033 singleton females (West et al. 2014).
50
The incidence of hirsutism was self-assessed by using a modified F&G score
sheet and a diagnosis of PCOS was enquired about via the questionnaire (answer
to the question: “Have you been diagnosed with PCOS by a physician?”). The
response rate to the 26-year-old’s questionnaire was 50.4%. The women with both
hirsutism (F&G score >7) and oligo- or amenorrhea according to the questionnaire
at the age of 26 years were considered as having PCOS according to both the NIH
(Zawadski & Dunaif 1992) and Rotterdam criteria (Rotterdam ESHRE/ASRMSponsored PCOS consensus workshop group 2004).
For measurement of serum AMH in Study II, we selected 400 subjects who had
responded to both questionnaires at the ages of 16 and 26 years. The study
population was split into testosterone quartiles and the study subjects were
randomly selected from each quartile (100 subjects from each testosterone quartile:
50 subjects with reported menstrual irregularities and 50 with normal menstrual
cycles at the age of 16 years) using a validated statistical method (SPSS software).
4.1.2 Nordic multicenter data (Studies III and IV)
In Study III, the participants in five Nordic PCOS studies were included (two
studies from Finland, two from Sweden and one from Norway), with a total of 681
women with PCOS (age range 18–59 years) and 230 healthy control women (age
range 19–62 years) (Hudecova et al. 2011, Morin-Papunen et al. 2000, Piltonen et
al. 2004, Piltonen et al. 2012, Puurunen et al. 2009, Stener-Victorin et al. 2010,
Vanky et al. 2004).
In Study IV, three more study sites were included (one from each: Sweden,
Norway and Denmark), resulting in a total number of 1550 PCOS subjects and 447
control women (Glintborg et al. 2012, Nybacka et al. 2011).
The diagnosis of PCOS was assessed according to the Rotterdam criteria.
Diagnosis of PCOS in peri- and postmenopausal women was based on oligoamenorrhea combined with hyperandrogenism (either biochemical or clinical)
reported at reproductive age, thus retrospectively meeting the Rotterdam criteria.
The control population consisted of women with normal ovaries in ultrasonography
and absence of PCOS-related symptoms, e.g. oligo- or anovulation and/or hirsutism.
51
4.1.3 Methods
Anthropometric measurements
In all studies, body mass index (BMI) was calculated by dividing the body weight
in kg by the squared height in meters. Waist and hip circumferences (measured to
the nearest centimeter with a soft tape at the midway level between the lowest rib
margin and the iliac crest and at the widest part of the gluteal region) were assessed
and the waist-to-hip ratio (WHR) was calculated. In Study IV, only the waist
circumference was used.
Laboratory methods
Androgens. In Studies I–III, serum samples for assay of testosterone were analyzed
by using Agilent triple quadrupole 6410 LC/MS equipment with an electrospray
ionization source operating in positive-ion mode (Agilent Technologies,
Wilmington, DE, USA). Multiple reaction monitoring was used to quantify
testosterone by using trideuterated testosterone (d3-testosterone), with the
following transitions: m/z 289.2 to 97 and 289.2 to 109 for testosterone and 292.2
to 97 and 292.2 to 109 for d3-testosterone. The intra-assay CVs of the method were
5.3%, 1.6% and 1.2% for testosterone at 0.6, 6.6 and 27.7 nmol/L, respectively. The
interassay CVs were 5.3%, 4.2% and 1.0% for the respective concentrations. The
three supplement populations in Study IV (Glintborg et al. 2012, Nybacka et al.
2011) had T measured with accredited methods in their respective laboratories prior
to inclusion in the study.
Androstenedione and DHEAS in Study III were assayed in the laboratories of
the different study centers according to their routine methods (RIA, immunoassay
and mass spectrometry, Table 7). The reference ranges were very similar in all
laboratories, except for a lower upper limit in mass spectrometry-analyzed samples
for A4 (Table 6).
Anti-Müllerian hormone. Serum AMH concentrations were assayed in the
stored samples by AMH Gen II enzyme-linked immunosorbent assay (ELISA,
Beckman Coulter Inc. 250 S. Kraemer Blvd. Brea, CA 92821 U.S.A.) as previously
described (Kumar et al. 2010, Wallace et al. 2011). Intra- and interassay
coefficients of variation were 4.6 and 8.0% at a concentration of 0.05 ng/mL. An
AMH level of 1 ng/mL equals 7.14 pmol/L.
52
Glucose, insulin, lipids and oral glucose tolerance test. In Studies I and II
plasma glucose, total cholesterol, high-density lipoprotein cholesterol (HDLcholesterol), low-density lipoprotein cholesterol (LDL-cholesterol) and
triglycerides were determined by using an automatic chemical analyzer (Cobas
Integra 700, Roche Diagnostics, Switzerland). Serum insulin was analyzed by
radioimmunoassay (Pharmacia Diagnostics, Uppsala, Sweden), high sensitivity Creactive protein (hs-CRP) by an immunoenzymometric assay (Medix Biochemica,
Espoo, Finland) and SHBG by time-resolved fluoroimmunoassay (AutoDelfia,
PerkinElmer, Turku, Finland).
In Studies III and IV metabolic parameters were measured by means of
accredited methods prior to inclusion in the study. The methods are described in
the original articles (Glintborg et al. 2012, Hudecova et al. 2009, Morin-Papunen
et al. 2000, Nybacka et al. 2011, Piltonen et al. 2004, Piltonen et al. 2012, Puurunen
et al. 2009, Stener-Victorin et al. 2010, Vanky et al. 2004).
Oral glucose tolerance tests (OGTTs) were carried out via the routine protocol
recommended for women with PCOS, with a fasting component and a 75-gram
glucose intake (Bozdag & Yildiz 2013). Mean OGTT glucose and serum insulin
levels were calculated as the means of concentrations at different time points [(basal
+ 2-hour) / 2].
FAI and cFT. The methods used to obtain the free androgen index and
calculated free testosterone are described above (Table 3).
To quantify the degree of insulin sensitivity and insulin resistance, homeostasis
model assessment (HOMA-S and HOMA-IR) values were calculated using the
validated calculator available at http://www.dtu.ox.ac.uk.
53
Table 6. Analysis methods for androstenedione and dehydroepiandrosterone sulfate in
Study III.
Androstenedione
Population
Method
Dehydroepiandrosterone sulfate
Reference
Method
Reference
RIAa
1–14 µmol/L
1–14 µmol/L
ranges
Population 1
d
1.4–14.3
Chemiluminometric
1.4–14.3
Chemiluminometric
Immunoassay / RIA1
nmol/L
immunoassay
h
GC-MSb
1.6–5.7 nmol/L
LC-MS/MSc
1.5–7.2 µmol/L
i
RIAa
0.7–11.0
Immunoassay
0.9–11.6
(n = 321)
Population 2
nmol/L
e,f,g
(n = 228)
Population 3
ranges
RIAa
(n = 128)
Population 4
nmol/L
(n = 90)
Population 5
j
Not analyzed
-
µmol/L
Not analyzed
-
(n = 145)
a
Radioimmunoassay, bGas chromatography–mass spectrometry, cLiquid chromatography–tandem mass
spectrometry, dMorin-Papunen et al. 2000, ePuurunen et al. 2009, fPiltonen et al. 2004, gPiltonen et al.
2012, hStener-Victorin et al. 2010, iVanky et al. 2004 and jHudecova et al. 2009.
4.2
Data analysis and statistical methods
The Chi squared test was used to compare categorical variables between the study
groups. Comparisons of continuous variables were conducted by using Student’s ttest or the (nonparametric) Mann–Whitney test, depending on the distribution of
the variable.
Trends in biochemical variables were assessed by means of analysis of variance
(ANOVA) or the Kruskal–Wallis test when having more than two study groups. In
order to be able to adjust for confounding variables, ANOVA and analysis of
covariance (ANCOVA) were adopted when comparing, for example, AMH levels
between two groups.
If the distribution of a laboratory measurement was skewed, it was logarithmically
transformed to achieve normality. If the parameter was further skewed, the
(nonparametric) Kruskal–Wallis test was used.
Correlations were tested using Pearson's and Spearman's correlation analyses.
Partial correlation analysis was used to control for confounding variables. Linear
regression analysis was used to calculate percentage differences in biochemical
characteristics between the study groups and for adjustments. The results were
adjusted, if needed, for smoking, socioeconomic status (SES), alcohol consumption,
BMI, WHR and age at menarche.
54
Specific methods in different studies
Study I. The whole study population was stratified into FAI and BMI quartiles and
analysis of variance was used to assess the trends in biochemical variables across
these quartiles. Confidence intervals were calculated using a CIA computer
program (Gardner & Altman 1988).
Statistical analyses were performed by using SPSS 17.0 software (IBM Corp.,
Armonk, NY. IBM SPSS Statistics for Windows, Version 17.0. Released 2010.).
Study II. The study population of 400 girls was constituted from four quartiles
of 100 women according to their testosterone levels (described previously, see 4.1).
Confidence intervals for proportions were calculated using a CIA computer
program (Gardner & Altman 1988). Bias-corrected and accelerated 95%
confidence intervals of the correlation coefficients were calculated with 1000
bootstrap re-samples. We constructed receiver operating characteristic (ROC)
curves with AMH and testosterone concentrations at the age of 16 years, and for
women with PCOS at the age of 26 years. The cut-off values of AMH and
testosterone levels were selected to identify PCOS subjects with the best sensitivity
and specificity.
Statistical analyses were performed by using SPSS 18.0 software (IBM Corp.,
Armonk, NY. IBM SPSS Statistics for Windows, Version 18.0. Released 2010.).
Study III. The PCOS and control groups were divided into seven age groups:
18–24, 25–29, 30–34, 35–39, 40–44, 45–49 and over 50 years. Risk ratios were
estimated by using logistic regression analysis in which the dichotomy of the
variables was created with a specific cut-off concentration (with the best
combination of sensitivity and specificity) analyzed from the ROC curve for each
variable, if needed.
Statistical analyses were performed by using SPSS 19.0 software (IBM Corp.,
Armonk, NY. IBM SPSS Statistics for Windows, Version 19.0. Released 2010.).
Study IV. The PCOS population was divided into hyperandrogenic PCOS (with
biochemical and/or clinical hyperandrogenism, HA-PCOS) and normoandrogenic
PCOS (women with both oligo-amenorrhea and polycystic ovaries in
ultrasonography, but no hyperandrogenism, NA-PCOS) groups. The control
population consisted of women with normal ovaries in ultrasonography and
absence of PCOS-related symptoms, e.g. oligo- or anovulation and/or hirsutism
and/or elevated T levels. The PCOS and control populations were grouped
according to age as follows: < 30 years old, 30–39 years old and > 39 years old.
55
Statistical analyses were performed by using SPSS 21.0 and 22.0 software
(IBM Corp., Armonk, NY. IBM SPSS Statistics for Windows, Version 21.0/22.0.
Released 2010.).
56
57
elevated both ovarian and
adrenal androgens through life
with metabolic risk factors. AMH compared with controls. The
best predictors for PCOS are
cFT, A4 and FAI.
menstrual irregularities, but not
at age 16 correlates with
hirsutism and PCOS at age 26
and associates with subclinical
metabolic alterations in
adolescence.
Women with PCOS exhibit
AMH levels at age 16 correlate
marker for hyperandrogenaemia with hyperandrogenaemia and
Menstrual irregularity is a good
*PCOS was defined by Rotterdam criteria.
Main results
Hirsutism and PCOS at age 26
BMI and metabolic parameters
androgens Age and BMI
metabolic parameters at age 16 Estimated parameters for free
Hormone assays
metabolic alterations in PCOS.
role than HA as regards
seems to play a more adverse
controls. Abdominal obesity
menopause compared with
from early adulthood until
unfavorable metabolic profile
Women with PCOS have
components
Metabolic syndrome and its
BMI and metabolic parameters
Hormone assays
PCOS: 14-59 years
PCOS: 18-59 years
Hormone assays, AMH BMI and Hormone assays
Menstrual cycle
PCOS*, N=1550
Controls: 18-62 years
Main study parameters
Study IV
Controls, N=447
Controls: 19-62 years
~16 years
Age
~16 years and 26 years
Normal menstrual cycle, N=200 PCOS*, N=681
symptomatic girls, N=l739
Study III
Controls, N=230
No. of subjects
Study II
Menstrual irregularity, N=200
Study I
Symptomatic girls, N=709 Non-
Study details
Table 7. Subjects, main study parameters and main results of the studies.
58
5
Results
5.1
Menstrual disorders and hyperandrogenemia in adolescence
and early adulthood (Studies I and II)
5.1.1 Menstrual disorders
According to the questionnaire, 709 (29%) girls had menstrual disorders
(symptomatic subjects) and 1739 had regular periods (non-symptomatic subjects)
at the age of 16 years. The symptomatic girls were comparable to their nonsymptomatic counterparts as regards alcohol consumption, smoking habits and SES.
Menarcheal age was slightly higher in symptomatic (13.1 years) than in nonsymptomatic girls (12.9 years, P < 0.001). Gynecological age, defined as the
number of years since menarche, was calculated (based on the date given in the
questionnaire) to be 2.34 (SD 1.1) years in the symptomatic girls and 2.59 (SD 1.09)
years in the non-symptomatic girls (P < 0.001). The results were therefore adjusted
for gynecological age.
Symptomatic and non-symptomatic girls were similar as regards to BMI,
serum levels of insulin, hs-CRP and lipids (serum levels of total cholesterol, HDL,
LDL and triglycerides), plasma glucose and HOMA-S. Symptomatic girls
exhibited significantly higher serum concentrations of testosterone, lower serum
levels of SHBG and higher FAI values (Table 8).
Table 8. Hormone parameters in the girls with self-reported menstrual disorders.
Non-symptomatic girls
Parameters
N
Geometric mean
Symptomatic girls
N
(IQR)
Testosterone
1739
[nmol/L]
SHBG
1739
51.49
3.08
(2.15, 4.74)
1.65
0.010
(1.31, 2.07)
709
(36.60, 68.90)
1739
Geometric mean
(IQR)
709
(1.28, 2.02)
[nmol/L]
FAI
1.59
P-value
48.99
0.042
(35.28, 65.85)
709
3.38
0.002
(2.27, 5.18)
IQR; interquartile range from 25th to 75th percentile.
In the group of symptomatic girls 7.8% (55/709) exhibited testosterone levels
above the upper normal limit (95th percentile), i.e. 2.68 nmol/L in this population,
59
which was significantly higher than in non-symptomatic girls (4.9%, 86/1739, P =
0.007).
Adjustments for BMI or WHR (data not shown) did not alter the
aforementioned associations. After adjusting for BMI the percentage differences
between the symptomatic and non-symptomatic groups increased as regards the
FAI (from 9.5% to 11.3%), testosterone (from 4.2% to 4.5%) and SHBG (-4.8% to
-6.1%). Significant differences remained after additional adjustments for smoking,
SES and alcohol consumption. Further adjustment for menarcheal or gynecological
age did not change the results (data not shown). With a threshold level of 2.68
nmol/L for T, the sensitivity and specificity of being “symptomatic” as regards
prediction of hyperandrogenemia were 39% and 72%, respectively.
In the whole study population, 3.5% (86/2448) of the girls exhibited both
menstrual disorders and elevated serum testosterone levels fulfilling the Rotterdam
criteria for PCOS.
5.1.2 Effect of hyperandrogenemia
In the whole study population, after stratifying the FAI values into quartiles, there
was a significant linear trend in the higher FAI quartiles towards lower serum HDLcholesterol levels (from 1.48, 95% CI [1.46–1.51] to 1.41 [1.39–1.43] mmol/L, P
< 0.001). Moreover, serum triglyceride levels increased significantly from the
second (0.72 [0.70–0.74] mmol/L) and third FAI quartiles (0.72 [0.70–0.75]
mmol/L) to the fourth (0.77 [0.75–0.80 mmol/L], P = 0.009 and P = 0.032,
respectively).
5.1.3 Effect of weight
In the whole population and both study groups, there were significant correlations
between BMI (and WHR) and hyperandrogenemia (especially FAI) and metabolic
parameters (hs-CRP, lipids and indicators of insulin sensitivity).
After stratifying the population into BMI quartiles, there was a significant trend
towards higher FAI values in the higher BMI quartiles in the symptomatic as well
as in the non-symptomatic girls. The differences in FAI values between the
symptomatic and non-symptomatic girls were significant in the two highest BMI
quartiles. In the highest BMI quartile (25.4 kg/m2, range 22.5–46.9 kg/m2),
symptomatic girls had significantly higher BMI (P = 0.005) and FAI values (P =
0.002), lower serum levels of SHBG (P = 0.016) and HDL-cholesterol (P = 0.002),
60
higher serum levels of insulin (P = 0.031) and lower HOMA-S values (P = 0.027)
than the non-symptomatic girls. Levels of hs-CRP did not differ between the two
groups. There was a significant linear trend towards higher CRP levels from the
lowest to the highest BMI quartiles in the whole study population (from 0.16 [95%
CI 0.14–0.18] to 0.48 mg/L [0.43–0.54], P < 0.001) and in both the symptomatic
(from 0.15 [0.12–0.18] to 0.52 mg/L [0.42–0.65], P < 0.001) and non-symptomatic
girls (from 0.17 [0.14–0.19] to 0.47 mg/L [0.41–0.53], P < 0.001).
5.1.4 Anti-Müllerian hormone
There was a significant correlation between AMH and testosterone levels at the age
of 16 years (r = 0.36, P < 0.001). After stratifying into T quartiles, the correlation
was significant only in the highest T quartile (r = 0.33, 95% CI 0.01–0.56, P = 0.001)
and AMH levels increased significantly from the lowest towards the highest T
quartile (P-value for trend < 0.001).
AMH levels at the age of 16 years were significantly higher among girls with
menstrual disorders compared with girls with normal menstrual cycles. At the age
of 26 years, women with menstrual disorders and women without such disorders
did not differ as regards their levels of AMH at the age of 16 years. Within the
highest testosterone quartile, the women with menstrual disorders at the age of 26
years had higher AMH levels at the age of 16 years, compared with women with
normal menstrual cycles at the age of 26 years, even after adjusting for T and BMI
(Figure 7A).
Levels of AMH at the age of 16 years were higher in women with hirsutism at
the age of 26 years compared with non-hirsute women and higher in women with
PCOS at the age of 26 years compared with subjects without PCOS. (Figure 7B).
The sensitivity of serum AMH (cut-off level 22.5 pmol/L, ROC-curve analysis)
for predicting PCOS at the age of 26 years was 85.7% and specificity was 37.5%.
The addition of T into the model did not significantly improve the accuracy of the
test.
In the whole study population, there was a significant but weak correlation
between serum AMH levels and BMI at the age of 16 (r = 0.124, P = 0.013), but
the significance disappeared after adjusting for T. There were no significant
correlations between AMH levels and metabolic indices (WHR, plasma fasting
glucose, serum insulin, HOMA-S, hs-CRP, serum total cholesterol, HDLcholesterol, LDL-cholesterol and triglycerides).
61
Fig. 7A.
Fig. 7B.
Fig. 7. A: Anti-Müllerian hormone levels at the age of 16 years according to the
menstrual pattern in the whole study population at the age of 16 years and in the highest
T-quartile at the age of 26 years in women with menstrual irregularities and women with
normal menstruation. Fig. 7. B: AMH levels at the age of 16 years in women with
hirsutism or PCOS at the age of 26 years, and controls. The 95% confidence intervals
are shown as error bars.
5.2
Endocrine features of PCOS throughout life (Study III)
Serum androgen levels in women with PCOS decreased from the age of 18 towards
menopause and then increased again after the age of 50. In the control group, serum
levels of androgens and SHBG were more stable during ageing.
Correlations between age and endocrine parameters are shown in Table 9. After
adjustment for BMI, all correlations except that for SHBG remained significant in
women with PCOS, whereas the correlations between age and SHBG, cFT and FAI
disappeared in the control population.
62
Table 9. Correlations between age and hormone parameters in the women with PCOS
and the controls.
Parameters
T
Control
P-value
PCOS
P-value
N/A
NS
r = -0.237
P < 0.001*
SHBG
r = 0.141
P = 0.035
r = -0.100
P = 0.012
A4
r = -0.328
P < 0.001*
r = -0.241
P < 0.001*
DHEAS
r = -0.258
P = 0.010*
r = -0.233
P < 0.001*
FAI
r = 0.205
P = 0.006
r = -0.113
P = 0.004*
cFT
r = 0.192
P = 0.010
r = -0.169
P < 0.001*
* Statistical significance remained after BMI-adjustment
5.2.1 Testosterone, FAI and cFT
Serum testosterone concentrations were significantly higher in women with PCOS
compared with control women in the first four age groups (from 18 to 39 yrs, P <
0.001) and the significance remained after adjusting for BMI (P < 0.001). Both cFT
and FAI were higher in women with PCOS at ages between 18–44 years and > 50
yrs and the significance remained in the first five age groups after adjustment for
BMI, except for FAI in the age group of 30–34 years (Figure 8).
63
Fig. 8. Mean levels of androgen parameters in different age groups. Error bars represent
the standard deviation and only P-values after adjustment for BMI are shown in the
figure.
64
5.2.2 Adrenal androgens
Serum levels of A4 were significantly higher in women with PCOS compared with
controls at the age of 18–34 and in women aged 50 years or more, and the
differences remained after adjustment for BMI. Serum DHEAS levels were
significantly higher in women with PCOS between the ages 25 to 29 and 50 years
or more after adjustment for BMI. Levels of DHEAS were higher in the control
women than in the PCOS population between the ages 40 to 44, but the difference
became non-significant after adjustment for BMI (Figure 9).
Fig. 9. Mean concentrations of adrenal androgens in different age groups. Error bars
represent the standard deviation and only P-values after adjustment for BMI are shown
in the figure.
65
5.2.3 Sex hormone-binding globulin
Serum concentrations of SHBG were lower in women with PCOS in all age groups
except for that of 45–49 years. After adjusting for BMI the significance remained
at ages 25–29, 40–44 and 50 years or more (Figure 10).
Fig. 10. Mean concentrations of sex hormone-binding globulin in different age groups.
Error bars represent the standard deviation and only P-values after adjustment for BMI
are shown in the figure.
5.2.4 Menstrual pattern
The occurrence of oligo-amenorrhea decreased with age in the PCOS population
(odds ratio [OR] = 0.96, 95% confidence interval 0.94–0.98, P = 0.001), whereas it
increased in the control population (OR = 1.10, 95% confidence interval 1.03–1.18,
P = 0.007).
5.2.5 Evaluation of the risk of PCOS
According to the results of logistic regression analysis, the best predictive
parameters regarding the risk of PCOS were cFT, A4 and FAI (Table 10).
Calculated free testosterone exhibited the best estimated risk ratio, and the best
sensitivity and specificity values according to the ROC curve in the subpopulation
of women of less than 40 years. The results remained unchanged after exclusion of
66
subjects whose serum A4 levels were measured by GC-MS (N=128). Adjustment
for age and BMI strengthened the estimated risk ratio for A4 and to a lesser degree
for T in the subpopulation of women less than 40 years of age (Table 10).
Table 10. Estimated risk ratios (odds ratios) for PCOS as regards hormone parameters
with specific cut-off values.
Parameters in subpopulations
Crude
Adjusted b
(OR, 95% CI) a
(OR, 95% CI) a
5.05 (3.65–7.00)***
4.67 (3.27–6.66)***
2.62 (1.91–3.59)
***
1.53 (1.06–2.22)**
6.71 (4.76–9.46)
***
4.35 (3.01–6.29)***
Whole study population, N ≈ 856 (cut-off)
T (≥1.1 nmol/L)
SHBG (<46.0 nmol/L)
FAI (≥ 2.39)
cFT (≥0.014 nmol/L)
7.90 (5.41–11.55)
DHEAS (≥3.8 µmol/L)
A4 (≥9.7 nmol/L)
***
5.10 (3.40–7.64)***
3.80 (2.44–5.90)
***
2.57 (1.59–4.14)***
6.16 (4.20–9.05)
***
5.75 (3.74–8.84)***
Population under 40 years old, N ≈ 678 (cut-off)
6.34 (4.24–9.47)***
T (≥1.08 nmol/L)
SHBG (<46.8 nmol/L)
3.02 (2.04–4.46)
FAI (≥ 2.34)
9.22 (5.98–14.22)
cFT (≥0.014 nmol/L)
***
12.97 (7.75–21.70)
DHEAS (≥3.7 µmol/L)
3.57 (2.08–6.09)
A4 (≥10.7 nmol/L)
a
***
***
***
6.63 (4.27–10.30)
***
7.00 (4.54–10.78)***
1.59 (1.02–2.46)**
6.64 (4.18–10.56)***
10.18 (5.87–17.64)***
3.35 (1.92–5.87)***
7.95 (4.83–13.09)***
b
Confidence interval of 95%, Adjustment for BMI and age, * P-value < 0.05, ** P-value < 0.01, *** P-
value < 0.001
5.3
Metabolic features of PCOS throughout life (Study IV)
5.3.1 Comparison of women with PCOS and control women in the
whole population
The women with PCOS had higher BMI, serum levels of insulin (fasting and
OGTT), total cholesterol, LDL, triglycerides (TGs) and blood pressure, and lower
serum levels of HDL compared with the control population after adjustment for
BMI (Table 11).
67
Table 11. Anthropometric and metabolic parameters as mean values in PCOS and
control populations. Standard deviation in parentheses.
Metabolic parameters
N
Control population
N
PCOS
Age [yr]
447
33.5 (9.9)
1550
30.0 (7.2)
P-value
<0.001
BMI [kg/m2]
447
25.9 (5.4)
1497
29.2 (6.9)
<0.001
Waist [cm]
312
87.6 (14.6)
1204
92.9 (17.5)
<0.001
Fasting glucose [mmol/L]
376
5.1 (0.9)
1104
5.1 (0.6)
NS
Fasting insulin [mU/L]
372
7.4 (6.0)
1093
12.3 (11.1)
<0.001*
Total cholesterol [mmol/L]
364
4.6 (0.9)
982
4.8 (1.0)
0.003*
HDL [mmol/L]
346
1.5 (0.3)
960
1.4 (0.4)
<0.001*
LDL [mmol/L]
347
2.6 (0.8)
863
2.9 (0.9)
<0.001*
Triglycerides [mmol/L]
366
0.9 (0.5)
974
1.3 (0.8)
<0.001*
OGTT glucose 2h
140
5.0 (1.3)
681
5.8 (1.8)
<0.001
140
5.0 (0.8)
681
5.5 (1.1)
<0.001
[mmol/L]
OGTT mean glucose
[mmol/l]
OGTT insulin 2h [mU/l]
152
27.4 (20.5)
860
71.7 (73.9)
<0.001*
OGTT mean insulin
152
17.2 (12.0)
840
42.6 (41.6)
<0.001*
318
118.2 (15.6)
1276
123.43 (16.2)
<0.001*
318
74.3 (12.1)
1276
78.5 (12.1)
<0.001*
159
1.5 (3.0)
761
2.8 (3.8)
<0.001
[mU/l]
Systolic blood pressure
[mmHg]
Diastolic blood pressure
[mmHg]
hsCRP [mg/l]
*P-value < 0.05 after BMI-adjusment.
68
Table 12. Anthropometric
and
metabolic
parameters
as
mean
values
in
hyperandrogenic and normoandrogenic PCOS populations and in the control
population. Standard deviation in parentheses.
Metabolic parameters
N
Controls
N
447
33.5
684
NA-
N
PCOS
Age
[year]
BMI
(9.9)
447
[kg/m2]
Waist
[mmol/l]
Fasting insulin
312
[mmol/l]
HDL
[mmol/l]
LDL
[mmol/l]
Triglycerides
[mmol/l]
OGTT Glucose 2h
[mmol/l]
OGTT mean glucose
[mmol/l]
OGTT Insulin 2h
376
372
364
346
347
366
140
140
152
152
318
[mmHg]
a
544
4.6
(0.9)
1.5
(0.3)
2.6
(0.8)
0.9
(0.5)
5.0
(1.3)
5.0
0.8)
27.4
17.2
118
74
368
364
349
367
442
442
476
1.5
(3.0)
604
5.1
(0.6)
12.0
4.7
(0.9)
1.3
(0.5)
2.8
(0.8)
1.2
(0.7)
5.9
(1.7)
5.4
(1.0)
67.8
465
40.9
552
537
123
603
586
504
596
238
238
376
78
367
2.8
(3.6)
29.4
<0.001
<0.001
0.027
93.6
<0.001
<0.001
NSd
5.1
NS
NS
NS
<0.001d
<0.001d
NS
0.041
0.004d
NS
<0.001d
<0.001d
0.013
<0.001d
<0.001d
NS
<0.001d
<0.001d
NS
<0.001d
<0.001
NS
<0.001
<0.001
NS
<0.001d
<0.001d
NS
<0.001d
<0.001d
NS
<0.001d
<0.001d
NS
<0.001d
<0.001d
NS
<0.001
<0.001
NS
(0.6)
12.4
4.8
(1.0)
1.4
(0.4)
2.9
(0.9)
1.2
(0.8)
6.0
(1.9)
5.5
(1.1)
76.0
44.5
(43.3)
657
124
(17)
657
(12)
468
NS
(79.0)
(16)
605
<0.001
(10.7)
(40.1)
605
<0.001
(17.0)
(69.4)
(12)
159
92.1
30.0
(6.7)
(11.4)
(16)
Diastolic blood pressure 318
[mg/L]
7.4
542
(12.0)
[mmHg]
hsCRP
5.1
(0.9)
811
(17.8)
(20.5)
[mU/l]
Systolic blood pressure
590
(6.0)
[mU/l]
OGTT mean insulin
87.6
28.8
P-valueb P-valuec
(7.4)
(7.0)
(14.6)
[mU/l]
Total cholesterol
666
P-valuea
PCOS
842
(7.0)
(5.4)
[cm]
Fasting glucose
25 .9
29.9
HA-
79
(12)
291
2.9
(4.0)
P-value between control group and NA-PCOS group, bP-value between control group and HA-PCOS
group, cP-value between NA- and HA-PCOS groups. *P-value < 0.05 after adjustment for BMI and age.
69
5.3.2 Comparisons between HA-PCOS, NA-PCOS and control women
and the different age groups
5.3.3 Body mass index and waist circumference
The HA-PCOS group presented higher BMIs and waist circumferences compared
with both the NA-PCOS and control populations, and the NA-PCOS group had
higher BMIs compared with the control women (Table 12).
In age group analyses, both PCOS populations had higher BMIs and waist
circumferences compared with the control women in all age groups. In addition,
women in the HA-PCOS group had higher BMIs and waist circumferences
compared with those in the NA-PCOS group in the groups aged 30 years or more
(Figure 11).
Fig. 11. Mean body mass indices and waist circumferences in the different age groups
and in the different subpopulations of the study. Error bars represent standard deviation.
70
5.3.4 Glucose and insulin metabolism
In the whole population, both PCOS populations were more hyperinsulinemic
compared with control women, and women in the NA-PCOS group were more
glucose intolerant, with higher 2-h glucose levels compared with the control
population (Table 12).
Women in the NA-PCOS group under the age of 40 and those in the HA-PCOS
group over the age of 39 exhibited significantly higher fasting insulin levels
compared with the control population. In the groups older than 30 years, the 2-hour
insulin levels and mean insulin levels in OGTTs were higher in both PCOS
populations compared with control women (Figure 12). The 2-hour glucose and
mean glucose levels in OGTTs were statistically higher in both the HA-PCOS
group (6.1 mmol/L [SD 2.0] and 5.7 mmol/L [SD 1.2] respectively) and the NAPCOS group (6.0 mmol/L [SD 1.6] and 5.6 mmol/L [SD 1.0]) compared with
control subjects (4.5 mmol/L [SD 0.5] and 4.6 mmol/L [SD 0.4]) between the ages
of 30 to 39 years.
Fig. 12. Mean OGTT insulin levels in the different age groups and in the different
subpopulations of the study. Error bars represent standard deviation. Results are
adjusted for BMI.
71
5.3.5 Lipids
In the whole population, serum levels of total cholesterol were significantly higher
in the HA-PCOS group compared with controls. In addition, serum HDL levels
were lower, and LDL and TG levels were higher in both PCOS populations
compared with the control population (Table 12).
In those under 30 years of age, both PCOS subpopulations showed worse lipid
profiles compared with the control women. At ages over 39 years, women in the
HA-PCOS group had higher triglyceride levels compared with control women
(Figure 13).
Fig. 13. Mean total cholesterol and triglyceride levels in the different age groups and in
the different subpopulations of the study. Error bars represent standard deviation.
Results are adjusted for BMI.
72
Serum levels of HDL were lower and levels of LDL higher in both PCOS
populations when compared with the control women in the age group under 30
years. Serum levels of LDL were higher in the HA-PCOS group compared with
both NA-PCOS and control women in the age group over 39 years (Figure 14).
Fig. 14. Mean HDL and LDL levels in the different age groups and in the different
subpopulations of the study. Error bars represent standard deviation. Results are
adjusted for BMI.
73
5.3.6 Blood pressure
Both systolic and diastolic blood pressures were significantly higher in both PCOS
subpopulations compared with the control women (Table 12). In the age group
analyses, HA-PCOS-group women had higher systolic blood pressures when over
30 years of age and NA-PCOS-group women in all age groups compared with the
control women. Diastolic blood pressure was higher in HA-PCOS-group women
compared with control women in all age groups, whereas in NA-PCOS-group
women diastolic blood pressure was higher only in those under 39 years of age
compared with controls (Figure 15).
Fig. 15. Mean systolic and diastolic blood pressures in the different age groups and in
the different subpopulations of the study. Error bars represent standard deviation.
Results are adjusted for BMI.
74
5.3.7 High-sensitivity C-reactive protein
Levels of high-sensitivity CRP were higher in all women with PCOS as well as in
both PCOS subpopulations compared with the control women, but the difference
became nonsignificant after adjustment for BMI (Table 12).
5.3.8 Prevalence of metabolic syndrome
The prevalence of MetS (Rotterdam criteria) was two- to fivefold higher in the HAPCOS group compared with its prevalence in the control women in all age groups
and higher in the NA-PCOS group than in control women under 40 years of age.
The prevalence of MetS was similar in the control and NA-PCOS populations
compared with an approximately twofold higher prevalence in the HA-PCOS group
when over 39 years of age (Figure 16). The results were similar when using the
criteria of the International Diabetes Federation to define metabolic syndrome
(Alberti et al. 2005).
75
Fig. 16. The prevalence of metabolic syndrome and its components in the different age
groups and in the different subpopulations of the study. Results are adjusted for BMI.
76
6
Discussion
In the present study, the associations between early menstrual irregularities,
determined by asking a simple question in adolescence, serum AMH levels,
hyperandrogenism and other metabolic parameters in adolescence were evaluated.
Moreover, we wanted to clarify the value of early menstrual irregularities and the
reliability of serum levels of AMH in predicting the emergence of PCOS in later
life. Further, we investigated the value of serum androgen levels from early until
late adulthood in diagnosis of the syndrome in a large Nordic population of women
with PCOS. Lastly, we investigated the evolution of the syndrome with age and the
relationships between the typical hormonal and metabolic abnormalities related to
the syndrome, such as abdominal obesity, hyperandrogenism, insulin resistance,
impaired glucose tolerance, dyslipidemia and chronic inflammation as well as the
prevalence of metabolic syndrome.
6.1
Early prediction of PCOS
6.1.1 Menstrual irregularities
In our study population of Finnish adolescent girls, one third of the subjects
exhibited menstrual irregularities, confirming that these are common gynecological
problems in adolescents.
Irregular menstruation during reproductive life has been suggested to be
associated with metabolic risks, such as increased risks of CVD, CVD events and
even increased mortality postmenopausally (Shaw et al. 2008). Menstrual
irregularities in adolescence have also been linked to hyperandrogenism (Lewy et
al. 2001). We were able to confirm these findings, as menstrual irregularities at the
age of 16 years were associated with an increased prevalence of hyperandrogenism
as well as with a lower level of SHBG, despite the fact that the symptomatic girls
were leaner than their asymptomatic controls. Even though there were no other
significant associations between menstrual irregularities and metabolic disorders at
that age, the presence of lower SHBG levels in symptomatic girls is of interest, as
SHBG has been shown to be a good surrogate marker of insulin resistance (Kajaia
et al. 2007). Low serum SHBG levels have also been associated with an increased
risk of T2DM in later life (Wallace et al. 2013). Our finding may therefore allow
us to suggest that menstrual irregularities in adolescence may be linked to
77
metabolic risks later in life. The symptomatic girls, however, did not exhibit other
metabolic risks linked to PCOS, possibly because they were very young and clinical
manifestation of insulin resistance and metabolic disorders had not yet developed.
Of note, the symptomatic group most probably also included girls with mild cycle
irregularities, which may have decreased the differences between the two groups.
Nevertheless, the symptomatic girls in the highest BMI quartile were more
hyperandrogenic and insulin resistant compared with their leaner counterparts, and
the symptomatic girls were more hyperandrogenic and insulin resistant compared
with non-symptomatic girls. The most hyperandrogenic girls in the whole study
population already presented a more adverse lipid profile at a young age. Lastly,
adjustment for BMI even increased the differences between the two groups.
In our study, the mean gynecological age in the symptomatic girls with
menstrual irregularities was only 2.5 years, suggesting that evaluation of the
menstrual cycle might have taken place too soon after menarche. However, the
results did not change after adjusting for gynecological age, which may suggest
that the presence of menstrual disorders as early as two to three years after
menarche could be a hallmark of hyperandrogenism and hyperinsulinemia, both
typical disorders in cases of PCOS. Moreover, our results are consistent with earlier
findings showing that 62% of adolescent girls suffer from irregular menstruation
and 59% of such girls fulfilled PCOS criteria at the end of a 6-year follow-up period
(Wiksten-Almstromer et al. 2008). Of interest is the fact that the girls with
menstrual irregularities had entered menarche three months later than girls with
regular menses, which could indicate that menstrual irregularities already existed
at menarche, but this hypothesis remains to be confirmed. Inversely, in our study,
the presence of regular menses at the age of 16 years diminished the risk of HA
with a specificity of 72% and could be a predictor of a normal hormonal status later
in life.
Overall, our results confirm that the association between obesity,
hyperandrogenism and metabolic risks is already present in adolescence. More
specifically, we could show that menstrual irregularities, identified via a simple
question at the time of adolescence, represent a good marker of hyperandrogenemia
and may be a risk factor as regards later metabolic risks and development of PCOS
in adulthood. Inversely, the presence of regular menses at the age of 16 years could
predict a normal hormonal status later in life.
78
6.1.2 Anti-Müllerian hormone
In Study II we aimed to clarify whether the relationship between AMH,
hyperandrogenism, and menstrual and metabolic disorders already exists in
adolescence and what could be the value of AMH as a predictive tool regarding the
development of oligo-amenorrhea, hirsutism and PCOS later in early adulthood.
We observed a significant positive correlation between serum concentrations
of T and AMH as early as in adolescence, as well as higher AMH levels in girls
with mentrual irregularities at the age of 16 years, in line with previous results
showing a positive association between the length of menstrual cycles and serum
levels of AMH at the age of 20 years (Kristensen et al. 2012). Our results also
showed also that AMH levels were associated significantly with oligo- or
amenorrhoea and were a good indicator of these conditions in adolescence.
Moreover, retrospectively, the women reporting hirsutism at the age of 26 years
had higher levels of AMH at the age of 16 years compared with non-hirsute women
at the age of 26 years, and women with PCOS (either diagnosed or self-reported
symptoms according to the questionnaire) at the age of 26 years had higher serum
AMH levels in adolescence compared with healthy women. Serum AMH levels at
the age of 16 years, however, did not differ significantly between women with
oligo- or amenorrhea and women without these conditions at the age of 26 years.
We calculated the specificity and sensitivity of serum levels of AMH in relation
to the risk of developing PCOS in later life, and serum AMH levels over 22.5
pmol/L could identify adolescent girls at risk of PCOS in early adulthood with good
sensitivity (85.7%). However, the specificity of the test remained very weak. It
improved substantially when using a higher cut-off value of 42.8 pmol/L, but the
sensitivity dropped and the combination of levels of AMH and T did not improve
the accuracy of the test, in line with the results of some (Casadei et al. 2013), but
not all studies (Eilertsen et al. 2012). Previous studies have indicated that AMH
alone may not be good enough as a single screening or diagnostic tool for PCOS
(Singh & Singh 2015). Our study indicates that serum AMH levels in adolescence
do not represent a reliable tool to predict PCOS in later life.
In adolescence no association between AMH and metabolic risks has been
shown (Anderson et al. 2013, Lin et al. 2011), but contradictory results later in life
have also been published (Nardo et al. 2009, Park et al. 2010, Skalba et al. 2011).
We found a weak correlation between AMH levels and BMI at the age of 16 years,
but after adjustment for T the result became nonsignificant. In addition, there was
79
no significant correlation between serum levels of AMH and metabolic indices,
indicating that AMH is a poor marker of an adverse metabolic profile in
adolescence, in line with previous results (Anderson et al. 2013, Cui et al. 2016).
In conclusion, girls with oligo- or amenorrhea at the age of 16 years or
PCOS/isolated hirsutism at the age of 26 years already exhibited higher levels of
AMH in adolescence, suggesting that the association between elevated AMH levels
and symptoms of PCOS in adolescence persists into early adulthood. As a result of
good sensitivity, use of a sufficiently high cut-off value of AMH could predict
which adolescents are likely to develop PCOS in adulthood, enabling management
of this condition from an early age. However, because of its low specificity, it is
not an ideal diagnostic marker, and its routine use in clinical practice cannot at
present be recommended. Lastly, AMH is not a good marker of cardiovascular risk
factors in adolescence.
6.2
Lifelong hormonal and metabolic parameters
6.2.1 Androgen parameters in PCOS across life
The data from this multicenter collaborative work carried out in Nordic countries
showed that women with PCOS exhibited significantly elevated serum androgen
levels, of both ovarian and adrenal origin. The differences were especially
significant in early adulthood and in the pre- and postmenopausal periods, as shown
previously in other ethnic populations (Liang et al. 2011, Winters et al. 2000). This
finding confirms that young women suffer more often from adverse features related
to hyperandrogenism (hirsutism and acne) than they do in later life. Additionally,
mood disorders and decreased quality of life seem to be worse in early adulthood
in women with PCOS (Dokras 2012, Johnstone et al. 2012, Livadas et al. 2011).
These findings strengthen the importance of implementing active therapeutic
intervention among young women in order to improve these adverse manifestations
of hyperandrogenism, as well as quality of life.
Women with PCOS remained more hyperandrogenic until the age of 44 years
when compared with the control population. After the steady decrease in androgen
levels up to menopause, hyperandrogenism persisted or even worsened after
menopause in women with PCOS, in line with the results of previous studies
(Burger et al. 2000, Puurunen et al. 2009, Schmidt et al. 2011). Increases in the FAI
and cFT after menopause were most probably related to the decrease of SHBG
80
levels, which has been linked to increased weight, and impaired glucose and insulin
metabolism (Cao et al. 2013, Markopoulos et al. 2011, Puurunen et al. 2009,
Puurunen et al. 2011). Supporting this hypothesis, after adjustment for BMI, the
differences in FAI and cFT between women with PCOS and non-PCOS subjects
became nonsignificant. We observed a similar result regarding A4 and DHEAS,
thus further supporting the strong correlation between hyperandrogenism and
obesity after menopause.
When evaluating the value of different parameters for assessing androgenicity,
cFT, the FAI and A4 were found to be the best risk factors as regards the prediction
of PCOS, thus supporting previous postulations that the FAI or cFT rather than T
should be used when defining hyperandrogenism in clinical practice (Cho et al.
2008, Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group
2004, Vermeulen et al. 1999). Some previous investigators have suggested that
serum A4 levels are independent of BMI and that A4 is the most consistently
elevated androgen in women with PCOS (Overlie et al. 1999, Puurunen et al. 2011).
After adjusting for BMI, A4 showed the best estimated risk ratio in the whole
population, but in the subgroup of women under 40 years, cFT exhibited the best
predictive value.
In conclusion, this large Nordic multicenter collaborative study showed that
obesity and hyperandrogenemia are worse throughout reproductive life in women
with PCOS compared with healthy control women. Hyperandrogenemia improves
toward menopause in women with PCOS, but remains elevated after menopause.
According to the results of logistic regression analysis, cFT, the FAI, and serum A4
are the best discriminators of PCOS at all ages and may be used as additional tools
for diagnosing PCOS in large populations. Lastly, in this population, the use of cFT
instead of the FAI as a discriminator of PCOS seems to be more accurate in women
younger than 40 years of age, at least when using LC-MS/MS for the assay of T.
6.2.2 Metabolic risk factors in PCOS across life
The main finding in this large Nordic collaborative study was that the unfavorable
metabolic profile in women with PCOS, independently of hyperandrogenism, was
already present in early adulthood and extended until menopause.
Studies carried out to evaluate metabolic abnormalities in women with PCOS
have mostly involved small numbers of subjects of relatively narrow age range, and
the role of HA as a metabolic risk factor in PCOS is still under debate (Moran &
81
Teede 2009, O'Reilly et al. 2014, Randeva et al. 2012). Some data suggest that the
metabolic risks linked to PCOS are mainly related to obesity and insulin resistance
(Moran & Teede 2009, Tzeng et al. 2014), whereas others support the hypothesis
that the syndrome per se and especially HA, independently of BMI, are the most
important cardiovascular risks (Nisenblat & Norman 2009, Park et al. 2010, Piouka
et al. 2009). It was therefore interesting to compare women in the NA- and HAPCOS groups as regards their metabolic profiles in our large Nordic population,
from early adulthood until menopause.
The results of some previous studies have suggested that the subpopulation of
normoandrogenic oligo-amenorrheic women with PCOS does not differ from
healthy BMI-matched controls as regards metabolic risks and that this subgroup of
women may present a more favorable metabolic phenotype compared with their
hyperandrogenic counterparts (Barber et al. 2007). More specifically, differences
between non-hyperandrogenic women with PCOS and healthy women have been
suggested to be mostly associated with increased abdominal obesity (Moran &
Teede 2009). Our present results support this postulation, as the NA-PCOS
population presented as a distinguishable subgroup as regards metabolic risks; such
women were more abdominally obese, and, even after adjusting for BMI, more
glucose intolerant, insulin resistant, hyperinsulinemic and dyslipidemic, especially
in early adulthood, compared with control women.
As for the role of HA as a metabolic risk in PCOS, women in both the NA- and
HA-PCOS groups exhibited more adverse metabolic alterations throughout their
fertile lives compared with control women, without significant differences between
the two PCOS subpopulations, except for significantly greater BMIs and waist
circumferences among those in the HA-PCOS group over 30 years of age. Again,
our results confirm a predominant role of obesity as regards the severity of the
syndrome (Moran & Teede 2009). On the other hand, HA may have some
significance as a late metabolic risk, as the prevalence of MetS was significantly
greater in women in the HA-PCOS group compared with the NA-PCOS group and
the control population after the age of 40 years, which may be partly due to the
relatively low numbers of women in these groups. Indeed, women in the NA-PCOS
group and control women exhibited similar waist circumferences at this age, which
may have biased the results regarding the prevalence of MetS, but again supporting
the postulation that abdominal obesity is the primary determinant of metabolic
abnormalities in PCOS (Moran & Teede 2009). This finding, however, is also
consistent with previous observations linking hyperandrogenism to metabolic
disturbances and an elevated prevalence of MetS (Sung et al. 2014). In keeping
82
with this postulation, in our study population, serum levels of LDL were slightly
higher in the HA-PCOS versus the NA-PCOS group in those over 40 years of age.
Even though the isolated parameters of the metabolic risk profile did not
considerably differ between PCOS subpopulations, lifelong exposure to
hyperandrogenism may end up being an additional risk as regards MetS later in life.
Hyperandrogenism may also directly or indirectly influence metabolic
abnormalities and contribute to abdominal obesity.
Women in both the HA- and NA-PCOS groups already exhibited adverse
changes in their metabolic profiles and blood pressure (BP) at early ages,
independently of obesity. In particular, high levels of triglycerides have been shown
to represent an independent predictor of future risk of myocardial infarction, and
low levels of HDL reflect cardiovascular morbidity (Stampfer et al. 1996) and are
associated with elevated cardiovascular risks in general, especially in women
(Ulmer et al. 2005). The difference in BP levels (4–5mm Hg between the PCOS
and the control groups), although relatively modest, may have clinical significance,
since it was already present in the younger age group (under 30 years). Moreover,
it has been shown that an increase of BP of 1.5–2 mm Hg could have a large impact
on CVD risk at a population level, and reduction of mildly and/or moderately
elevated BP to normal levels has been associated with a reduced risk of CVD in
large placebo-controlled studies (Staessen et al. 1997). All these observations fit
with recent epidemiologic data indicating that women with PCOS may have an
increased risk of ischemic heart and cerebrovascular diseases (Hart & Doherty
2015).
In summary, the results highlight the importance of screening for PCOS and
overweight/obesity in early adulthood in order to tailor treatment and intervention
protocols and reduce the risks of future CVD and T2DM. However, more research
is still needed to provide a greater understanding of the interaction of
hyperandrogenism, insulin resistance and abdominal adiposity in PCOS (Tzeng et
al. 2014), and only long-term follow-up will reveal whether the metabolic risk
factors linked to PCOS, and more specifically to HA, translate into later
cardiovascular morbidity and mortality.
83
6.3
Strengths and limitations of the studies
6.3.1 Studies I and II (NFBC-1986)
The strength of Studies I and II is the large, stable, population-based birth cohort.
The answer and participation rates were good, so that generalization of the results
to the Finnish population is appropriate. In addition, the questionnaires and clinical
evaluations were standardized, which further increases the reliability of our
findings.
In these studies serum levels of testosterone were determined by LC-MS/MS,
the most modern and accurate method for T analysis. The method for the
determination of AMH levels has been challenged recently, but the Gen II assay
used in the present study has been considered to be the most reliable method (Singh
& Singh 2015). Moreover, the determinations in our study were carried out at the
same time and with the same assay, thus decreasing the risks of inaccuracy of the
results.
One limitation in the studies is the questionnaire-based data for some
parameters (self-reported menstrual irregularities and hirsutism). As the study was
not designed to diagnose PCOS, neither ultrasonography nor clinical evaluations
of hirsutism were carried out, which can be considered as a limitation. However, in
a similar Northern Finland cohort, the Northern Finland Birth Cohort 1966, it has
previously been shown that self-reported menstrual irregularities and hirsutism
correlate with an endocrine profile typical of PCOS and that two simple questions
on menstrual irregularities and hirsutism are sufficient to detect women with PCOS
in the general population (Taponen et al. 2003).
The girls were not screened for late-onset adrenal hyperplasia, but the
prevalence of this disorder (under 1/15 000) in our country is considerably lower
than that reported for other populations (Jaaskelainen et al. 1997).
An important pitfall is that at the time of performance of our study, AMH
assays lacked an agreed international standard and therefore the concentrations and
cut-off points were method-dependent. The different immunoassays in use have
been compared and the results have been comparable, but different threshold values
should be used for automatic and manual assays (Pigny et al. 2016). According to
a recent review on the diagnostic significance of AMH in PCOS (Singh & Singh
2015), the Gen II assay used in the present study appears to be the most reliable
method for AMH measurement. The development of new methods for the
measurement of AMH will reveal whether the observed associations can be
84
confirmed. Importantly, the development of an international standard is needed in
the future.
6.3.2 Studies III and IV (Nordic population)
The strength of this Nordic collaborative study is the large number of subjects from
a homogeneous Caucasian population. Moreover, the study population includes
both women with PCOS and control women with a wide age range from
adolescence until the postmenopausal period. As in the NFBC-1986 study, the gold
standard method for measuring T in women, LC-MS/MS, was used in Study III and
in 52% of the subjects in Study IV.
One important limitation is the cross-sectional study design, which can cause
bias due to heterogeneity in the characteristics of the participants in the different
age groups. Furthermore, the control group was smaller than the PCOS group, as
not all trials involved in this collaborative study included a control population. The
methods for laboratory analyses varied between subpopulations at different study
sites. However, all were carried out according to accredited methods, and specific
reference ranges in the laboratories were used in the definition of biochemical
hyperandrogenism. The upper limit for T to define biochemical hyperandrogenism
did not take into account the physiological decline in T serum levels observed with
age. Thus, some hyperandrogenic women older than 39 years may have been
classified as normoandrogenic, which could narrow the differences between the
HA- and NA-PCOS groups. In Study IV, we could not clearly distinguish
premenopausal, menopausal and postmenopausal women, as the number of women
over 50 years of age was too small to obtain reliable results when comparing
different age groups.
85
86
7
1.
2.
3.
4.
5.
6.
Conclusions and future plans
Menstrual irregularities, enquired about in adolescence by means of one simple
question, represent a good marker of hyperandrogenemia and may predict the
development of PCOS. The simultaneous presence of menstrual irregularities
and overweight is associated with the worst metabolic risk profile at a young
age. Menstrual irregularities should be noted early, as an association with
hyperandrogenism, obesity and adverse metabolic risk factors is already
present in adolescence.
At the age of 16 years serum AMH levels correlate with those of T, and
menstrual irregularities. In addition, elevated AMH levels in adolescence are
associated with hirsutism and PCOS in early adulthood. However, AMH levels
in adolescence do not represent a reliable tool to predict PCOS in later life.
AMH is not a good marker of metabolic abnormalities in adolescence. In
adolescence, establishing the presence of menstrual irregularities by means of
a simple question, and measuring AMH levels, especially in hyperandrogenic
girls, could help to predict the later development of PCOS and, further, allow
the possibility to take preventive actions to avoid its development, which
would decrease the burden of the syndrome at both an individual level and
among the general public.
Women with PCOS have elevated androgen levels and are more obese
compared with control women throughout their fertile lives.
Hyperandrogenism seems to ease with age in women with PCOS, but androgen
levels remain elevated after menopause.
The FAI, cFT, and serum A4 are the best discriminators of PCOS at all ages
and may be used as additional tools for diagnosing PCOS in large populations.
Nordic women with PCOS present a worsened metabolic profile throughout
their fertile lives. Abdominal obesity seems to exert a more predominant role
than hyperandrogenism as regards the severity of the syndrome during the
reproductive period. However, lifelong exposure to elevated androgen levels
might result in a worsened metabolic burden in women with PCOS aged 40
years or more.
In order to tailor preventive actions and treatment of PCOS, age, androgenic status,
obesity and clinical features of the disorder should all be taken into account.
87
7.1.1 Future plans
A follow-up investigation including a new questionnaire as well as clinical
examinations is planned among the NFBC-1986 in 2018–2019, when the women
will be 31–32 years of age. One of the most important challenges will be to check
reproductive function as well as the hormonal and metabolic indices in these
women and to investigate the relationships with their hormonal and metabolic
indices at the age of 16 years. More specifically, we will investigate how menstrual
irregularities, metabolic and hormonal parameters and serum levels of AMH at the
age of 16 years correlate with reproductive and metabolic health at the age of 31–
32 years. Determination of serum levels of AMH at the age of 31–32 years will also
clarify the role and significance of AMH as a diagnostic tool for PCOS and as a
hypothetical indicator of hormonal and metabolic disorders in this population. This
follow-up study could help to clarify how the endocrine and metabolic profiles in
adolescence translate into reproductive and metabolic health in adulthood.
In the Nordic collaboration project, we are planning to carry out studies
focusing more specifically on chronic inflammation, and screening of dyslipidemia
and glucose metabolism disorders in women with PCOS. It would also be of
interest to investigate, in this same population, the prevalence of hirsutism and its
relationship with hyperandrogenemia. In addition, it would be of great interest to
measure serum AMH levels at different ages in this large population, and to
evaluate the correlations between AMH, and hormonal and metabolic parameters.
Further, in the future extended collaboration with other researchers in the field
could be considered, and a large European cohort of women with PCOS would
make it possible to reach deeper understanding of the syndrome.
88
References
Abbott DH, Barnett DK, Bruns CM & Dumesic DA (2005) Androgen excess fetal
programming of female reproduction: a developmental aetiology for polycystic ovary
syndrome? Hum Reprod Update 11(4): 357-374.
Abbott DH, Dumesic DA, Eisner JR, Colman RJ & Kemnitz JW (1998) Insights into the
development of polycystic ovary syndrome (PCOS) from studies of prenatally
androgenized female rhesus monkeys. Trends Endocrinol Metab 9(2): 62-67.
Abbott DH, Dumesic DA & Franks S (2002) Developmental origin of polycystic ovary
syndrome - a hypothesis. J Endocrinol 174(1): 1-5.
Abraham GE, Maroulis GB & Marshall JR (1974) Evaluation of ovulation and corpus
luteum function using measurements of plasma progesterone. Obstet Gynecol 44(4):
522-525.
Adams J, Polson DW & Franks S (1986) Prevalence of polycystic ovaries in women with
anovulation and idiopathic hirsutism. Br Med J (Clin Res Ed) 293(6543): 355-359.
Ader M, Stefanovski D, Richey JM, Kim SP, Kolka CM, Ionut V, Kabir M & Bergman RN
(2014) Failure of homeostatic model assessment of insulin resistance to detect marked
diet-induced insulin resistance in dogs. Diabetes 63(6): 1914-1919.
Aiyappan SK, Karpagam B, Vadanika V, Chidambaram PK, Vinayagam S & Saravanan KC
(2016) Age-Related Normogram for Ovarian Antral Follicle Count in Women with
Polycystic Ovary Syndrome and Comparison with Age Matched Controls Using
Magnetic Resonance Imaging. J Clin Diagn Res 10(1): TC11-3.
Ala-Fossi SL, Maenpaa J, Aine R & Punnonen R (1998) Ovarian testosterone secretion
during perimenopause. Maturitas 29(3): 239-245.
Alberti KG, Zimmet P, Shaw J & IDF Epidemiology Task Force Consensus Group (2005)
The metabolic syndrome--a new worldwide definition. Lancet 366(9491): 1059-1062.
Alsamarai S, Adams JM, Murphy MK, Post MD, Hayden DL, Hall JE & Welt CK (2009)
Criteria for polycystic ovarian morphology in polycystic ovary syndrome as a function
of age. J Clin Endocrinol Metab 94(12): 4961-4970.
Amsterdam ESHRE/ASRM-Sponsored 3rd PCOS Consensus Workshop Group (2012)
Consensus on women's health aspects of polycystic ovary syndrome (PCOS). Hum
Reprod 27(1): 14-24.
Anderson EL, Fraser A, McNally W, Sattar N, Lashen H, Fleming R, Nelson SM & Lawlor
DA (2013) Anti-mullerian hormone is not associated with cardiometabolic risk factors
in adolescent females. PLoS One 8(5): e64510.
Apter D (1980) Serum steroids and pituitary hormones in female puberty: a partly
longitudinal study. Clin Endocrinol (Oxf) 12(2): 107-120.
Apter D & Vihko R (1983) Early menarche, a risk factor for breast cancer, indicates early
onset of ovulatory cycles. J Clin Endocrinol Metab 57(1): 82-86.
Aydin Y, Hassa H, Burkankulu D, Arslantas D, Sayiner D & Ozerdogan N (2015) What is
the Risk of Metabolic Syndrome in Adolescents with Normal BMI who have Polycystic
Ovary Syndrome? J Pediatr Adolesc Gynecol 28(4): 271-274.
89
Azziz R, Carmina E, Dewailly D, Diamanti-Kandarakis E, Escobar-Morreale HF, Futterweit
W, Janssen OE, Legro RS, Norman RJ, Taylor AE, Witchel SF & Androgen Excess
Society (2006) Positions statement: criteria for defining polycystic ovary syndrome as
a predominantly hyperandrogenic syndrome: an Androgen Excess Society guideline. J
Clin Endocrinol Metab 91(11): 4237-4245.
Azziz R, Carmina E, Dewailly D, Diamanti-Kandarakis E, Escobar-Morreale HF, Futterweit
W, Janssen OE, Legro RS, Norman RJ, Taylor AE, Witchel SF & Task Force on the
Phenotype of the Polycystic Ovary Syndrome of The Androgen Excess and PCOS
Society (2009) The Androgen Excess and PCOS Society criteria for the polycystic
ovary syndrome: the complete task force report. Fertil Steril 91(2): 456-488.
Azziz R, Sanchez LA, Knochenhauer ES, Moran C, Lazenby J, Stephens KC, Taylor K &
Boots LR (2004) Androgen excess in women: experience with over 1000 consecutive
patients. J Clin Endocrinol Metab 89(2): 453-462.
Baillargeon JP, Jakubowicz DJ, Iuorno MJ, Jakubowicz S & Nestler JE (2004) Effects of
metformin and rosiglitazone, alone and in combination, in nonobese women with
polycystic ovary syndrome and normal indices of insulin sensitivity. Fertil Steril 82(4):
893-902.
Baptiste CG, Battista MC, Trottier A & Baillargeon JP (2010) Insulin and hyperandrogenism
in women with polycystic ovary syndrome. J Steroid Biochem Mol Biol 122(1-3): 4252.
Barber TM, Wass JA, McCarthy MI & Franks S (2007) Metabolic characteristics of women
with polycystic ovaries and oligo-amenorrhoea but normal androgen levels:
implications for the management of polycystic ovary syndrome. Clin Endocrinol (Oxf)
66(4): 513-517.
Barker DJ, Osmond C & Law CM (1989) The intrauterine and early postnatal origins of
cardiovascular disease and chronic bronchitis. J Epidemiol Community Health 43(3):
237-240.
Bergman RN, Hope ID, Yang YJ, Watanabe RM, Meador MA, Youn JH & Ader M (1989)
Assessment of insulin sensitivity in vivo: a critical review. Diabetes Metab Rev 5(5):
411-429.
Bili H, Laven J, Imani B, Eijkemans MJ & Fauser BC (2001) Age-related differences in
features associated with polycystic ovary syndrome in normogonadotrophic oligoamenorrhoeic infertile women of reproductive years. Eur J Endocrinol 145(6): 749-755.
Blank SK, Helm KD, McCartney CR & Marshall JC (2008) Polycystic ovary syndrome in
adolescence. Ann N Y Acad Sci 1135: 76-84.
Boomsma CM, Eijkemans MJ, Hughes EG, Visser GH, Fauser BC & Macklon NS (2006)
A meta-analysis of pregnancy outcomes in women with polycystic ovary syndrome.
Hum Reprod Update 12(6): 673-683.
Bozdag G & Yildiz BO (2013) Interventions for the metabolic dysfunction in polycystic
ovary syndrome. Steroids 78(8): 777-781.
Broekmans FJ & Fauser BC (2006) Diagnostic criteria for polycystic ovarian syndrome.
Endocrine 30(1): 3-11.
90
Bui HN, Sluss PM, Blincko S, Knol DL, Blankenstein MA & Heijboer AC (2013) Dynamics
of serum testosterone during the menstrual cycle evaluated by daily measurements with
an ID-LC-MS/MS method and a 2nd generation automated immunoassay. Steroids
78(1): 96-101.
Burger HG (2002) Androgen production in women. Fertil Steril 77 Suppl 4: S3-5.
Burger HG, Dudley EC, Cui J, Dennerstein L & Hopper JL (2000) A prospective
longitudinal study of serum testosterone, dehydroepiandrosterone sulfate, and sex
hormone-binding globulin levels through the menopause transition. J Clin Endocrinol
Metab 85(8): 2832-2838.
Burger HG, Hale GE, Robertson DM & Dennerstein L (2007) A review of hormonal changes
during the menopausal transition: focus on findings from the Melbourne Women's
Midlife Health Project. Hum Reprod Update 13(6): 559-565.
Burgers JA, Fong SL, Louwers YV, Valkenburg O, de Jong FH, Fauser BC & Laven JS
(2010) Oligoovulatory and anovulatory cycles in women with polycystic ovary
syndrome (PCOS): what's the difference? J Clin Endocrinol Metab 95(12): E485-9.
Cao YK, Zhang SF, Zou SE, Xia X & Xu LN (2013) Study on relationship between
endogenous androgens and insulin resistance at the different stages of postmenopause.
Zhonghua Fu Chan Ke Za Zhi 48(10): 740-744.
Carmina E, Campagna AM, Mansuet P, Vitale G, Kort D & Lobo R (2012) Does the level
of serum antimullerian hormone predict ovulatory function in women with polycystic
ovary syndrome with aging? Fertil Steril 98(4): 1043-1046.
Carmina E, Koyama T, Chang L, Stanczyk FZ & Lobo RA (1992) Does ethnicity influence
the prevalence of adrenal hyperandrogenism and insulin resistance in polycystic ovary
syndrome? Am J Obstet Gynecol 167(6): 1807-1812.
Carmina E, Oberfield SE & Lobo RA (2010) The diagnosis of polycystic ovary syndrome
in adolescents. Am J Obstet Gynecol 203(3): 201.e1-201.e5.
Carmina E, Rosato F, Janni A, Rizzo M & Longo RA (2006) Extensive clinical experience:
relative prevalence of different androgen excess disorders in 950 women referred
because of clinical hyperandrogenism. J Clin Endocrinol Metab 91(1): 2-6.
Carroll J, Saxena R & Welt CK (2012) Environmental and genetic factors influence age at
menarche in women with polycystic ovary syndrome. J Pediatr Endocrinol Metab 25(56): 459-466.
Casadei L, Madrigale A, Puca F, Manicuti C, Emidi E, Piccione E & Dewailly D (2013) The
role of serum anti-Mullerian hormone (AMH) in the hormonal diagnosis of polycystic
ovary syndrome. Gynecol Endocrinol 29(6): 545-550.
Catteau-Jonard S & Dewailly D (2013) Pathophysiology of polycystic ovary syndrome: the
role of hyperandrogenism. Front Horm Res 40: 22-27.
Cedars MI, Steingold KA, de Ziegler D, Lapolt PS, Chang RJ & Judd HL (1992) Long-term
administration of gonadotropin-releasing hormone agonist and dexamethasone:
assessment of the adrenal role in ovarian dysfunction. Fertil Steril 57(3): 495-500.
91
Chen MJ, Yang WS, Chen CL, Wu MY, Yang YS & Ho HN (2008) The relationship
between anti-Mullerian hormone, androgen and insulin resistance on the number of
antral follicles in women with polycystic ovary syndrome. Hum Reprod 23(4): 952-957.
Cho LW, Kilpatrick ES, Jayagopal V, Diver MJ & Atkin SL (2008) Biological variation of
total testosterone, free androgen index and bioavailable testosterone in polycystic
ovarian syndrome: implications for identifying hyperandrogenaemia. Clin Endocrinol
(Oxf) 68(3): 390-394.
Cimino I, Casoni F, Liu X, Messina A, Parkash J, Jamin SP, Catteau-Jonard S, Collier F,
Baroncini M, Dewailly D, Pigny P, Prescott M, Campbell R, Herbison AE, Prevot V &
Giacobini P (2016) Novel role for anti-Mullerian hormone in the regulation of GnRH
neuron excitability and hormone secretion. Nat Commun 7: 10055.
Cinar N, Harmanci A, Aksoy DY, Aydin K & Yildiz BO (2012) Adrenocortical steroid
response to ACTH in different phenotypes of non-obese polycystic ovary syndrome. J
Ovarian Res 5(1): 42-2215-5-42.
Couzinet B, Meduri G, Lecce MG, Young J, Brailly S, Loosfelt H, Milgrom E & Schaison
G (2001) The postmenopausal ovary is not a major androgen-producing gland. J Clin
Endocrinol Metab 86(10): 5060-5066.
Cui L, Qin Y, Gao X, Lu J, Geng L, Ding L, Qu Z, Zhang X & Chen ZJ (2016) Antimullerian
hormone: correlation with age and androgenic and metabolic factors in women from
birth to postmenopause. Fertil Steril 105(2): 481-485.e1.
Daan NM, Louwers YV, Koster MP, Eijkemans MJ, de Rijke YB, Lentjes EW, Fauser BC
& Laven JS (2014) Cardiovascular and metabolic profiles amongst different polycystic
ovary syndrome phenotypes: who is really at risk? Fertil Steril 102(5): 1444-1451.e3.
Danesh J, Wheeler JG, Hirschfield GM, Eda S, Eiriksdottir G, Rumley A, Lowe GD, Pepys
MB & Gudnason V (2004) C-reactive protein and other circulating markers of
inflammation in the prediction of coronary heart disease. N Engl J Med 350(14): 13871397.
Davison SL, Bell R, Donath S, Montalto JG & Davis SR (2005) Androgen levels in adult
females: changes with age, menopause, and oophorectomy. J Clin Endocrinol Metab
90(7): 3847-3853.
de Kat AC, Broekmans FJ, Laven JS & van der Schouw YT (2015) Anti-Mullerian Hormone
as a marker of ovarian reserve in relation to cardio-metabolic health: a narrative review.
Maturitas 80(3): 251-257.
DeFronzo RA, Tobin JD & Andres R (1979) Glucose clamp technique: a method for
quantifying insulin secretion and resistance. Am J Physiol 237(3): E214-23.
Dewailly D, Gronier H, Poncelet E, Robin G, Leroy M, Pigny P, Duhamel A & CatteauJonard S (2011) Diagnosis of polycystic ovary syndrome (PCOS): revisiting the
threshold values of follicle count on ultrasound and of the serum AMH level for the
definition of polycystic ovaries. Hum Reprod 26(11): 3123-3129.
Dewailly D, Lujan ME, Carmina E, Cedars MI, Laven J, Norman RJ & Escobar-Morreale
HF (2014) Definition and significance of polycystic ovarian morphology: a task force
report from the Androgen Excess and Polycystic Ovary Syndrome Society. Hum
Reprod Update 20(3): 334-352.
92
Dewailly D, Pigny P, Soudan B, Catteau-Jonard S, Decanter C, Poncelet E & Duhamel A
(2010) Reconciling the definitions of polycystic ovary syndrome: the ovarian follicle
number and serum anti-Mullerian hormone concentrations aggregate with the markers
of hyperandrogenism. J Clin Endocrinol Metab 95(9): 4399-4405.
Diamanti-Kandarakis E, Kouli C, Alexandraki K & Spina G (2004) Failure of mathematical
indices to accurately assess insulin resistance in lean, overweight, or obese women with
polycystic ovary syndrome. J Clin Endocrinol Metab 89(3): 1273-1276.
Diamanti-Kandarakis E, Mitrakou A, Hennes MM, Platanissiotis D, Kaklas N, Spina J,
Georgiadou E, Hoffmann RG, Kissebah AH & Raptis S (1995) Insulin sensitivity and
antiandrogenic therapy in women with polycystic ovary syndrome. Metabolism 44(4):
525-531.
Diamanti-Kandarakis E & Panidis D (2007) Unravelling the phenotypic map of polycystic
ovary syndrome (PCOS): a prospective study of 634 women with PCOS. Clin
Endocrinol (Oxf) 67(5): 735-742.
Dokras A (2012) Mood and anxiety disorders in women with PCOS. Steroids 77(4): 338341.
Dramusic V, Goh VH, Rajan U, Wong YC & Ratnam SS (1997) Clinical, endocrinologic,
and ultrasonographic features of polycystic ovary syndrome in Singaporean adolescents.
J Pediatr Adolesc Gynecol 10(3): 125-132.
Dumesic DA, Oberfield SE, Stener-Victorin E, Marshall JC, Laven JS & Legro RS (2015)
Scientific Statement on the Diagnostic Criteria, Epidemiology, Pathophysiology, and
Molecular Genetics of Polycystic Ovary Syndrome. Endocr Rev 36(5): 487-525.
Dunaif A, Segal KR, Shelley DR, Green G, Dobrjansky A & Licholai T (1992) Evidence for
distinctive and intrinsic defects in insulin action in polycystic ovary syndrome. Diabetes
41(10): 1257-1266.
Durlinger AL, Kramer P, Karels B, de Jong FH, Uilenbroek JT, Grootegoed JA & Themmen
AP (1999) Control of primordial follicle recruitment by anti-Mullerian hormone in the
mouse ovary. Endocrinology 140(12): 5789-5796.
Durlinger AL, Visser JA & Themmen AP (2002) Regulation of ovarian function: the role of
anti-Mullerian hormone. Reproduction 124(5): 601-609.
Ehrmann DA, Barnes RB, Rosenfield RL, Cavaghan MK & Imperial J (1999) Prevalence of
impaired glucose tolerance and diabetes in women with polycystic ovary syndrome.
Diabetes Care 22(1): 141-146.
Eilertsen TB, Vanky E & Carlsen SM (2012) Anti-Mullerian hormone in the diagnosis of
polycystic ovary syndrome: can morphologic description be replaced? Hum Reprod
27(8): 2494-2502.
Elting MW, Korsen TJ, Bezemer PD & Schoemaker J (2001) Prevalence of diabetes mellitus,
hypertension and cardiac complaints in a follow-up study of a Dutch PCOS population.
Hum Reprod 16(3): 556-560.
Elting MW, Korsen TJ, Rekers-Mombarg LT & Schoemaker J (2000) Women with
polycystic ovary syndrome gain regular menstrual cycles when ageing. Hum Reprod
15(1): 24-28.
93
Escobar-Morreale HF, Luque-Ramirez M & Gonzalez F (2011) Circulating inflammatory
markers in polycystic ovary syndrome: a systematic review and metaanalysis. Fertil
Steril 95(3): 1048-58.e1-2.
Essah PA, Nestler JE & Carmina E (2008) Differences in dyslipidemia between American
and Italian women with polycystic ovary syndrome. J Endocrinol Invest 31(1): 35-41.
Ewing JA & Rouse BA (1978) Hirsutism, race and testosterone levels: comparison of East
Asians and Euroamericans. Hum Biol 50(2): 209-215.
Fanchin R, Schonauer LM, Righini C, Guibourdenche J, Frydman R & Taieb J (2003) Serum
anti-Mullerian hormone is more strongly related to ovarian follicular status than serum
inhibin B, estradiol, FSH and LH on day 3. Hum Reprod 18(2): 323-327.
Ferriman D & Gallway JD (1961) Clinical assessment of body hair growth in women. J Clin
Endocrinol Metab 21: 1440-1447.
Franks S (1995) Polycystic ovary syndrome. N Engl J Med 333(13): 853-861.
Fraser IS & Kovacs G (2004) Current recommendations for the diagnostic evaluation and
follow-up of patients presenting with symptomatic polycystic ovary syndrome. Best
Pract Res Clin Obstet Gynaecol 18(5): 813-823.
Fruzzetti F, Perini D, Lazzarini V, Parrini D & Genazzani AR (2009) Hyperandrogenemia
influences the prevalence of the metabolic syndrome abnormalities in adolescents with
the polycystic ovary syndrome. Gynecol Endocrinol 25(5): 335-343.
Gardner MJ & Altman DG (1988) Estimating with confidence. Br Med J (Clin Res Ed)
296(6631): 1210-1211.
Gilling-Smith C, Willis DS, Beard RW & Franks S (1994) Hypersecretion of
androstenedione by isolated thecal cells from polycystic ovaries. J Clin Endocrinol
Metab 79(4): 1158-1165.
Glintborg D, Hass Rubin K, Nybo M, Abrahamsen B & Andersen M (2015a) Morbidity and
medicine prescriptions in a nationwide Danish population of patients diagnosed with
polycystic ovary syndrome. Eur J Endocrinol 172(5): 627-638.
Glintborg D, Mumm H, Ravn P & Andersen M (2012) Age associated differences in
prevalence of individual rotterdam criteria and metabolic risk factors during
reproductive age in 446 caucasian women with polycystic ovary syndrome. Horm
Metab Res 44(9): 694-698.
Glintborg D, Sidelmann JJ, Altinok ML, Mumm H & Andersen M (2015b) Increased
thrombin generation in women with polycystic ovary syndrome: A pilot study on the
effect of metformin and oral contraceptives. Metabolism 64(10): 1272-1278.
Glueck CJ, Morrison JA, Goldenberg N & Wang P (2009) Coronary heart disease risk
factors in adult premenopausal white women with polycystic ovary syndrome compared
with a healthy female population. Metabolism 58(5): 714-721.
GOLDZIEHER JW & AXELROD LR (1963) Clinical and Biochemical Features of
Polycystic Ovarian Disease. Fertil Steril 14: 631-653.
Goodarzi MO, Carmina E & Azziz R (2014) DHEA, DHEAS and PCOS. J Steroid Biochem
Mol Biol .
94
Gosman GG, King WC, Schrope B, Steffen KJ, Strain GW, Courcoulas AP, Flum DR,
Pender JR & Simhan HN (2010) Reproductive health of women electing bariatric
surgery. Fertil Steril 94(4): 1426-1431.
Gruijters MJ, Visser JA, Durlinger AL & Themmen AP (2003) Anti-Mullerian hormone and
its role in ovarian function. Mol Cell Endocrinol 211(1-2): 85-90.
Hammond GL & Bocchinfuso WP (1996) Sex hormone-binding globulin: gene organization
and structure/function analyses. Horm Res 45(3-5): 197-201.
Han AR, Kim HO, Cha SW, Park CW, Kim JY, Yang KM, Song IO, Koong MK & Kang
IS (2011) Adverse pregnancy outcomes with assisted reproductive technology in nonobese women with polycystic ovary syndrome: a case-control study. Clin Exp Reprod
Med 38(2): 103-108.
Handelsman DJ & Wartofsky L (2013) Requirement for mass spectrometry sex steroid
assays in the Journal of Clinical Endocrinology and Metabolism. J Clin Endocrinol
Metab 98(10): 3971-3973.
Hanson MA & Gluckman PD (2008) Developmental origins of health and disease: new
insights. Basic Clin Pharmacol Toxicol 102(2): 90-93.
Harrison CL, Lombard CB, Moran LJ & Teede HJ (2011) Exercise therapy in polycystic
ovary syndrome: a systematic review. Hum Reprod Update 17(2): 171-183.
Hart R & Doherty DA (2015) The potential implications of a PCOS diagnosis on a woman's
long-term health using data linkage. J Clin Endocrinol Metab 100(3): 911-919.
Hatch R, Rosenfield RL, Kim MH & Tredway D (1981) Hirsutism: implications, etiology,
and management. Am J Obstet Gynecol 140(7): 815-830.
Hillman JK, Johnson LN, Limaye M, Feldman RA, Sammel M & Dokras A (2014) Black
women with polycystic ovary syndrome (PCOS) have increased risk for metabolic
syndrome and cardiovascular disease compared with white women with PCOS
[corrected. Fertil Steril 101(2): 530-535.
Hines M, Golombok S, Rust J, Johnston KJ, Golding J & Avon Longitudinal Study of
Parents and Children Study Team (2002) Testosterone during pregnancy and gender
role behavior of preschool children: a longitudinal, population study. Child Dev 73(6):
1678-1687.
Hoeger KM & Oberfield SE (2012) Do women with PCOS have a unique predisposition to
obesity? Fertil Steril 97(1): 13-17.
Hoffman LK & Ehrmann DA (2008) Cardiometabolic features of polycystic ovary syndrome.
Nat Clin Pract Endocrinol Metab 4(4): 215-222.
Hucking K, Watanabe RM, Stefanovski D & Bergman RN (2008) OGTT-derived measures
of insulin sensitivity are confounded by factors other than insulin sensitivity itself.
Obesity (Silver Spring) 16(8): 1938-1945.
Hudecova M, Holte J, Moby L, Olovsson M, Stridsberg M, Larsson A, Berglund L, Berne
C & Sundstrom Poromaa I (2011) Androgen levels, insulin sensitivity, and early insulin
response in women with polycystic ovary syndrome: a long-term follow-up study. Fertil
Steril 95(3): 1146-1148.
95
Hudecova M, Holte J, Olovsson M & Sundstrom Poromaa I (2009) Long-term follow-up of
patients with polycystic ovary syndrome: reproductive outcome and ovarian reserve.
Hum Reprod 24(5): 1176-1183.
Hughes CL,Jr, Wall LL & Creasman WT (1991) Reproductive hormone levels in
gynecologic oncology patients undergoing surgical castration after spontaneous
menopause. Gynecol Oncol 40(1): 42-45.
Hull MG (1987) Epidemiology of infertility and polycystic ovarian disease:
endocrinological and demographic studies. Gynecol Endocrinol 1(3): 235-245.
Ibanez L, Dimartino-Nardi J, Potau N & Saenger P (2000) Premature adrenarche--normal
variant or forerunner of adult disease? Endocr Rev 21(6): 671-696.
Ibanez L, Jaramillo A, Enriquez G, Miro E, Lopez-Bermejo A, Dunger D & de Zegher F
(2007) Polycystic ovaries after precocious pubarche: relation to prenatal growth. Hum
Reprod 22(2): 395-400.
Ibanez L, Potau N, Chacon P, Pascual C & Carrascosa A (1998) Hyperinsulinaemia,
dyslipaemia and cardiovascular risk in girls with a history of premature pubarche.
Diabetologia 41(9): 1057-1063.
Jaaskelainen J, Levo A, Voutilainen R & Partanen J (1997) Population-wide evaluation of
disease manifestation in relation to molecular genotype in steroid 21-hydroxylase
(CYP21) deficiency: good correlation in a well defined population. J Clin Endocrinol
Metab 82(10): 3293-3297.
Jakimiuk AJ, Weitsman SR, Navab A & Magoffin DA (2001) Luteinizing hormone receptor,
steroidogenesis acute regulatory protein, and steroidogenic enzyme messenger
ribonucleic acids are overexpressed in thecal and granulosa cells from polycystic
ovaries. J Clin Endocrinol Metab 86(3): 1318-1323.
Janse F, Eijkemans MJ, Goverde AJ, Lentjes EG, Hoek A, Lambalk CB, Hickey TE, Fauser
BC & Norman RJ (2011) Assessment of androgen concentration in women: liquid
chromatography-tandem mass spectrometry and extraction RIA show comparable
results. Eur J Endocrinol 165(6): 925-933.
Johnstone EB, Davis G, Zane LT, Cedars MI & Huddleston HG (2012) Age-related
differences in the reproductive and metabolic implications of polycystic ovarian
syndrome: findings in an obese, United States population. Gynecol Endocrinol 28(10):
819-822.
Johnstone EB, Rosen MP, Neril R, Trevithick D, Sternfeld B, Murphy R, Addauan-Andersen
C, McConnell D, Pera RR & Cedars MI (2010) The polycystic ovary post-rotterdam: a
common, age-dependent finding in ovulatory women without metabolic significance. J
Clin Endocrinol Metab 95(11): 4965-4972.
Joseph DR (1994) Structure, function, and regulation of androgen-binding protein/sex
hormone-binding globulin. Vitam Horm 49: 197-280.
Judd HL, Judd GE, Lucas WE & Yen SS (1974) Endocrine function of the postmenopausal
ovary: concentration of androgens and estrogens in ovarian and peripheral vein blood.
J Clin Endocrinol Metab 39(6): 1020-1024.
96
Kajaia N, Binder H, Dittrich R, Oppelt PG, Flor B, Cupisti S, Beckmann MW & Mueller A
(2007) Low sex hormone-binding globulin as a predictive marker for insulin resistance
in women with hyperandrogenic syndrome. Eur J Endocrinol 157(4): 499-507.
Kazemi H, Ramezani Tehrani F, Minooee S, Khalili D & Azizi F (2015) Women selfperception of excess hair growth, as a predictor of clinical hirsutism: a population-based
study. J Endocrinol Invest 38(8): 923-928.
Kent SC & Legro RS (2002) Polycystic ovary syndrome in adolescents. Adolesc Med 13(1):
73-88, vi.
Kristensen SL, Ramlau-Hansen CH, Andersen CY, Ernst E, Olsen SF, Bonde JP, Vested A
& Toft G (2012) The association between circulating levels of antimullerian hormone
and follicle number, androgens, and menstrual cycle characteristics in young women.
Fertil Steril 97(3): 779-785.
La Marca A, Broekmans FJ, Volpe A, Fauser BC, Macklon NS & ESHRE Special Interest
Group for Reproductive Endocrinology--AMH Round Table (2009) Anti-Mullerian
hormone (AMH): what do we still need to know? Hum Reprod 24(9): 2264-2275.
Laitinen J, Taponen S, Martikainen H, Pouta A, Millwood I, Hartikainen AL, Ruokonen A,
Sovio U, McCarthy MI, Franks S & Jarvelin MR (2003) Body size from birth to
adulthood as a predictor of self-reported polycystic ovary syndrome symptoms. Int J
Obes Relat Metab Disord 27(6): 710-715.
Legro RS (2012) Obesity and PCOS: implications for diagnosis and treatment. Semin
Reprod Med 30(6): 496-506.
Legro RS, Arslanian SA, Ehrmann DA, Hoeger KM, Murad MH, Pasquali R, Welt CK &
Endocrine Society (2013) Diagnosis and treatment of polycystic ovary syndrome: an
Endocrine Society clinical practice guideline. J Clin Endocrinol Metab 98(12): 45654592.
Legro RS, Kunselman AR, Dodson WC & Dunaif A (1999) 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. J Clin Endocrinol
Metab 84(1): 165-169.
Legro RS, Roller RL, Dodson WC, Stetter CM, Kunselman AR & Dunaif A (2010a)
Associations of birthweight and gestational age with reproductive and metabolic
phenotypes in women with polycystic ovarian syndrome and their first-degree relatives.
J Clin Endocrinol Metab 95(2): 789-799.
Legro RS, Schlaff WD, Diamond MP, Coutifaris C, Casson PR, Brzyski RG, Christman GM,
Trussell JC, Krawetz SA, Snyder PJ, Ohl D, Carson SA, Steinkampf MP, Carr BR,
McGovern PG, Cataldo NA, Gosman GG, Nestler JE, Myers ER, Santoro N, Eisenberg
E, Zhang M, Zhang H & Reproductive Medicine Network (2010b) Total testosterone
assays in women with polycystic ovary syndrome: precision and correlation with
hirsutism. J Clin Endocrinol Metab 95(12): 5305-5313.
Lewy VD, Danadian K, Witchel SF & Arslanian S (2001) Early metabolic abnormalities in
adolescent girls with polycystic ovarian syndrome. J Pediatr 138(1): 38-44.
97
Liang SJ, Hsu CS, Tzeng CR, Chen CH & Hsu MI (2011) Clinical and biochemical
presentation of polycystic ovary syndrome in women between the ages of 20 and 40.
Hum Reprod .
Lin YH, Chiu WC, Wu CH, Tzeng CR, Hsu CS & Hsu MI (2011) Antimullerian hormone
and polycystic ovary syndrome. Fertil Steril 96(1): 230-235.
Livadas S, Chaskou S, Kandaraki AA, Skourletos G, Economou F, Christou M, Boutzios G,
Karachalios A, Zerva A, Xyrafis X, Christakou C, Pighou AK & Diamanti-Kandarakis
E (2011) Anxiety is associated with hormonal and metabolic profile in women with
polycystic ovarian syndrome. Clin Endocrinol (Oxf) 75(5): 698-703.
Lobo RA, Paul WL & Goebelsmann U (1981) Dehydroepiandrosterone sulfate as an
indicator of adrenal androgen function. Obstet Gynecol 57(1): 69-73.
Longcope C (1986) Adrenal and gonadal androgen secretion in normal females. Clin
Endocrinol Metab 15(2): 213-228.
March WA, Moore VM, Willson KJ, Phillips DI, Norman RJ & Davies MJ (2010) The
prevalence of polycystic ovary syndrome in a community sample assessed under
contrasting diagnostic criteria. Hum Reprod 25(2): 544-551.
Markopoulos MC, Kassi E, Alexandraki KI, Mastorakos G & Kaltsas G (2015)
Hyperandrogenism after menopause. Eur J Endocrinol 172(2): R79-91.
Markopoulos MC, Rizos D, Valsamakis G, Deligeoroglou E, Grigoriou O, Chrousos GP,
Creatsas G & Mastorakos G (2011) Hyperandrogenism in women with polycystic ovary
syndrome persists after menopause. J Clin Endocrinol Metab 96(3): 623-631.
Marshburn PB & Carr BR (1995) Hirsutism and virilization. A systematic approach to
benign and potentially serious causes. Postgrad Med 97(1): 99-102, 105-6.
Matthews KA, Crawford SL, Chae CU, Everson-Rose SA, Sowers MF, Sternfeld B &
Sutton-Tyrrell K (2009) Are changes in cardiovascular disease risk factors in midlife
women due to chronological aging or to the menopausal transition? J Am Coll Cardiol
54(25): 2366-2373.
Mendez-Sanchez N, Arrese M, Zamora-Valdes D & Uribe M (2007) Current concepts in the
pathogenesis of nonalcoholic fatty liver disease. Liver Int 27(4): 423-433.
Miller KK, Rosner W, Lee H, Hier J, Sesmilo G, Schoenfeld D, Neubauer G & Klibanski A
(2004) Measurement of free testosterone in normal women and women with androgen
deficiency: comparison of methods. J Clin Endocrinol Metab 89(2): 525-533.
Morales AJ, Laughlin GA, Butzow T, Maheshwari H, Baumann G & Yen SS (1996) Insulin,
somatotropic, and luteinizing hormone axes in lean and obese women with polycystic
ovary syndrome: common and distinct features. J Clin Endocrinol Metab 81(8): 28542864.
Moran C, Garcia-Hernandez E, Barahona E, Gonzalez S & Bermudez JA (2003)
Relationship between insulin resistance and gonadotropin dissociation in obese and
nonobese women with polycystic ovary syndrome. Fertil Steril 80(6): 1466-1472.
Moran L & Teede H (2009) Metabolic features of the reproductive phenotypes of polycystic
ovary syndrome. Hum Reprod Update 15(4): 477-488.
98
Moran LJ, Misso ML, Wild RA & Norman RJ (2010) Impaired glucose tolerance, type 2
diabetes and metabolic syndrome in polycystic ovary syndrome: a systematic review
and meta-analysis. Hum Reprod Update 16(4): 347-363.
Moran LJ, Teede HJ, Noakes M, Clifton PM, Norman RJ & Wittert GA (2013) Sex hormone
binding globulin, but not testosterone, is associated with the metabolic syndrome in
overweight and obese women with polycystic ovary syndrome. J Endocrinol Invest
36(11): 1004-1010.
Morikawa M, Derynck R, Miyazono K. (2016) TGF-β and the TGF-β Family: ContextDependent Roles in Cell and Tissue Physiology. Cold Spring Harb Perspect Biol. 2016
May 2;8(5).
Morin-Papunen LC, Vauhkonen I, Koivunen RM, Ruokonen A & Tapanainen JS (2000)
Insulin sensitivity, insulin secretion, and metabolic and hormonal parameters in healthy
women and women with polycystic ovarian syndrome. Hum Reprod 15(6): 1266-1274.
Mortensen M, Ehrmann DA, Littlejohn E & Rosenfield RL (2009) Asymptomatic volunteers
with a polycystic ovary are a functionally distinct but heterogeneous population. J Clin
Endocrinol Metab 94(5): 1579-1586.
Nardo LG, Yates AP, Roberts SA, Pemberton P & Laing I (2009) The relationships between
AMH, androgens, insulin resistance and basal ovarian follicular status in non-obese
subfertile women with and without polycystic ovary syndrome. Hum Reprod 24(11):
2917-2923.
Nduwayo L, Despert F, Lecomte C & Lecomte P (1992) Primary amenorrhea revealing
micropolycystic ovary syndrome. Presse Med 21(23): 1060-1063.
Nelson LR & Bulun SE (2001) Estrogen production and action. J Am Acad Dermatol 45(3
Suppl): S116-24.
Nisenblat V & Norman RJ (2009) Androgens and polycystic ovary syndrome. Curr Opin
Endocrinol Diabetes Obes 16(3): 224-231.
Nybacka A, Carlstrom K, Stahle A, Nyren S, Hellstrom PM & Hirschberg AL (2011)
Randomized comparison of the influence of dietary management and/or physical
exercise on ovarian function and metabolic parameters in overweight women with
polycystic ovary syndrome. Fertil Steril 96(6): 1508-1513.
O'Reilly MW, Taylor AE, Crabtree NJ, Hughes BA, Capper F, Crowley RK, Stewart PM,
Tomlinson JW & Arlt W (2014) Hyperandrogenemia predicts metabolic phenotype in
polycystic ovary syndrome: the utility of serum androstenedione. J Clin Endocrinol
Metab 99(3): 1027-1036.
Orio F,Jr, Palomba S, Cascella T, De Simone B, Di Biase S, Russo T, Labella D, Zullo F,
Lombardi G & Colao A (2004) Early impairment of endothelial structure and function
in young normal-weight women with polycystic ovary syndrome. J Clin Endocrinol
Metab 89(9): 4588-4593.
Overlie I, Moen MH, Morkrid L, Skjaeraasen JS & Holte A (1999) The endocrine transition
around menopause--a five years prospective study with profiles of gonadotropines,
estrogens, androgens and SHBG among healthy women. Acta Obstet Gynecol Scand
78(7): 642-647.
99
Pabon JE, Li X, Lei ZM, Sanfilippo JS, Yussman MA & Rao CV (1996) Novel presence of
luteinizing hormone/chorionic gonadotropin receptors in human adrenal glands. J Clin
Endocrinol Metab 81(6): 2397-2400.
Panidis D, Tziomalos K, Chatzis P, Papadakis E, Delkos D, Tsourdi EA, Kandaraki EA &
Katsikis I (2013) Association between menstrual cycle irregularities and endocrine and
metabolic characteristics of the polycystic ovary syndrome. Eur J Endocrinol 168(2):
145-152.
Park HT, Cho GJ, Ahn KH, Shin JH, Kim YT, Hur JY, Kim SH, Lee KW & Kim T (2010)
Association of insulin resistance with anti-Mullerian hormone levels in women without
polycystic ovary syndrome (PCOS). Clin Endocrinol (Oxf) 72(1): 26-31.
Pasquali R, Casimirri F, De Iasio R, Mesini P, Boschi S, Chierici R, Flamia R, Biscotti M &
Vicennati V (1995) Insulin regulates testosterone and sex hormone-binding globulin
concentrations in adult normal weight and obese men. J Clin Endocrinol Metab 80(2):
654-658.
Pawelczak M, Kenigsberg L, Milla S, Liu YH & Shah B (2012) Elevated serum antiMullerian hormone in adolescents with polycystic ovary syndrome: relationship to
ultrasound features. J Pediatr Endocrinol Metab 25(9-10): 983-989.
Payne AH & Hales DB (2004) Overview of steroidogenic enzymes in the pathway from
cholesterol to active steroid hormones. Endocr Rev 25(6): 947-970.
Peter A, Kantartzis K, Machann J, Schick F, Staiger H, Machicao F, Schleicher E, Fritsche
A, Haring HU & Stefan N (2010) Relationships of circulating sex hormone-binding
globulin with metabolic traits in humans. Diabetes 59(12): 3167-3173.
Pigny P, Gorisse E, Ghulam A, Robin G, Catteau-Jonard S, Duhamel A & Dewailly D (2016)
Comparative assessment of five serum antimullerian hormone assays for the diagnosis
of polycystic ovary syndrome. Fertil Steril 105(4): 1063-1069.e3.
Pigny P, Jonard S, Robert Y & Dewailly D (2006) Serum anti-Mullerian hormone as a
surrogate for antral follicle count for definition of the polycystic ovary syndrome. J Clin
Endocrinol Metab 91(3): 941-945.
Pigny P, Merlen E, Robert Y, Cortet-Rudelli C, Decanter C, Jonard S & Dewailly D (2003)
Elevated serum level of anti-mullerian hormone in patients with polycystic ovary
syndrome: relationship to the ovarian follicle excess and to the follicular arrest. J Clin
Endocrinol Metab 88(12): 5957-5962.
Piltonen T, Koivunen R, Morin-Papunen L, Ruokonen A, Huhtaniemi IT & Tapanainen JS
(2002) Ovarian and adrenal steroid production: regulatory role of LH/HCG. Hum
Reprod 17(3): 620-624.
Piltonen T, Koivunen R, Perheentupa A, Morin-Papunen L, Ruokonen A & Tapanainen JS
(2004) Ovarian age-related responsiveness to human chorionic gonadotropin in women
with polycystic ovary syndrome. J Clin Endocrinol Metab 89(8): 3769-3775.
Piltonen T, Koivunen R, Ruokonen A & Tapanainen JS (2003) Ovarian age-related
responsiveness to human chorionic gonadotropin. J Clin Endocrinol Metab 88(7): 33273332.
100
Piltonen T, Morin-Papunen L, Koivunen R, Perheentupa A, Ruokonen A & Tapanainen JS
(2005) Serum anti-Mullerian hormone levels remain high until late reproductive age
and decrease during metformin therapy in women with polycystic ovary syndrome.
Hum Reprod 20(7): 1820-1826.
Piltonen T, Puurunen J, Hedberg P, Ruokonen A, Mutt SJ, Herzig KH, Nissinen A, MorinPapunen L & Tapanainen JS (2012) Oral, transdermal and vaginal combined
contraceptives induce an increase in markers of chronic inflammation and impair
insulin sensitivity in young healthy normal-weight women: a randomized study. Hum
Reprod 27(10): 3046-3056.
Pinola P, Piltonen TT, Puurunen J, Vanky E, Sundstrom-Poromaa I, Stener-Victorin E,
Ruokonen A, Puukka K, Tapanainen JS & Morin-Papunen LC (2015) Androgen Profile
Through Life in Women With Polycystic Ovary Syndrome: A Nordic Multicenter
Collaboration Study. J Clin Endocrinol Metab 100(9): 3400-3407.
Piouka A, Farmakiotis D, Katsikis I, Macut D, Gerou S & Panidis D (2009) Anti-Mullerian
hormone levels reflect severity of PCOS but are negatively influenced by obesity:
relationship with increased luteinizing hormone levels. Am J Physiol Endocrinol Metab
296(2): E238-43.
Plymate SR, Jones RE, Matej LA & Friedl KE (1988) Regulation of sex hormone binding
globulin (SHBG) production in Hep G2 cells by insulin. Steroids 52(4): 339-340.
Polotsky AJ, Allshouse AA, Crawford SL, Harlow SD, Khalil N, Kazlauskaite R, Santoro
N & Legro RS (2014) Hyperandrogenic oligomenorrhea and metabolic risks across
menopausal transition. J Clin Endocrinol Metab 99(6): 2120-2127.
Polson DW, Adams J, Wadsworth J & Franks S (1988) Polycystic ovaries--a common
finding in normal women. Lancet 1(8590): 870-872.
Prelevic GM, Beljic T, Balint-Peric L & Ginsburg J (1995) Cardiac flow velocity in women
with the polycystic ovary syndrome. Clin Endocrinol (Oxf) 43(6): 677-681.
Purifoy FE, Koopmans LH & Tatum RW (1980) Steroid hormones and aging: free
testosterone, testosterone and androstenedione in normal females aged 20-87 years.
Hum Biol 52(2): 181-191.
Puurunen J, Piltonen T, Jaakkola P, Ruokonen A, Morin-Papunen L & Tapanainen JS (2009)
Adrenal androgen production capacity remains high up to menopause in women with
polycystic ovary syndrome. J Clin Endocrinol Metab 94(6): 1973-1978.
Puurunen J, Piltonen T, Morin-Papunen L, Perheentupa A, Jarvela I, Ruokonen A &
Tapanainen JS (2011) Unfavorable hormonal, metabolic, and inflammatory alterations
persist after menopause in women with PCOS. J Clin Endocrinol Metab 96(6): 18271834.
Rachmiel M, Kives S, Atenafu E & Hamilton J (2008) Primary amenorrhea as a
manifestation of polycystic ovarian syndrome in adolescents: a unique subgroup? Arch
Pediatr Adolesc Med 162(6): 521-525.
101
Rajpert-De Meyts E, Jorgensen N, Graem N, Muller J, Cate RL & Skakkebaek NE (1999)
Expression of anti-Mullerian hormone during normal and pathological gonadal
development: association with differentiation of Sertoli and granulosa cells. J Clin
Endocrinol Metab 84(10): 3836-3844.
Randeva HS, Tan BK, Weickert MO, Lois K, Nestler JE, Sattar N & Lehnert H (2012)
Cardiometabolic aspects of the polycystic ovary syndrome. Endocr Rev 33(5): 812-841.
Rebora A (2004) Pathogenesis of androgenetic alopecia. J Am Acad Dermatol 50(5): 777779.
Rojas J, Chavez M, Olivar L, Rojas M, Morillo J, Mejias J, Calvo M & Bermudez V (2014)
Polycystic ovary syndrome, insulin resistance, and obesity: navigating the
pathophysiologic labyrinth. Int J Reprod Med 2014: 719050.
Rosenfield RL, Lipton RB & Drum ML (2009) Thelarche, pubarche, and menarche
attainment in children with normal and elevated body mass index. Pediatrics 123(1):
84-88.
Rosner W, Auchus RJ, Azziz R, Sluss PM & Raff H (2007) Position statement: Utility,
limitations, and pitfalls in measuring testosterone: an Endocrine Society position
statement. J Clin Endocrinol Metab 92(2): 405-413.
Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group (2004) Revised
2003 consensus on diagnostic criteria and long-term health risks related to polycystic
ovary syndrome (PCOS). Hum Reprod 19(1): 41-47.
Ruutiainen K, Erkkola R, Gronroos MA & Kaihola HL (1988) Androgen parameters in
hirsute women: correlations with body mass index and age. Fertil Steril 50(2): 255-259.
Salameh WA, Redor-Goldman MM, Clarke NJ, Mathur R, Azziz R & Reitz RE (2014)
Specificity and predictive value of circulating testosterone assessed by tandem mass
spectrometry for the diagnosis of polycystic ovary syndrome by the National Institutes
of Health 1990 criteria. Fertil Steril 101(4): 1135-1141.e2.
Schmidt J, Brannstrom M, Landin-Wilhelmsen K & Dahlgren E (2011) Reproductive
hormone levels and anthropometry in postmenopausal women with polycystic ovary
syndrome (PCOS): a 21-year follow-up study of women diagnosed with PCOS around
50 years ago and their age-matched controls. J Clin Endocrinol Metab 96(7): 2178-2185.
Schmidt TH & Shinkai K (2015) Evidence-based approach to cutaneous hyperandrogenism
in women. J Am Acad Dermatol 73(4): 672-690.
Shaw LJ, Bairey Merz CN, Azziz R, Stanczyk FZ, Sopko G, Braunstein GD, Kelsey SF, Kip
KE, Cooper-Dehoff RM, Johnson BD, Vaccarino V, Reis SE, Bittner V, Hodgson TK,
Rogers W & Pepine CJ (2008) Postmenopausal women with a history of irregular
menses and elevated androgen measurements at high risk for worsening cardiovascular
event-free survival: results from the National Institutes of Health--National Heart, Lung,
and Blood Institute sponsored Women's Ischemia Syndrome Evaluation. J Clin
Endocrinol Metab 93(4): 1276-1284.
Shea JL, Wongt PY & Chen Y (2014) Free testosterone: clinical utility and important
analytical aspects of measurement. Adv Clin Chem 63: 59-84.
102
Simo R, Saez-Lopez C, Barbosa-Desongles A, Hernandez C & Selva DM (2015) Novel
insights in SHBG regulation and clinical implications. Trends Endocrinol Metab 26(7):
376-383.
Singh AK & Singh R (2015) Can anti-Mullerian hormone replace ultrasonographic
evaluation in polycystic ovary syndrome? A review of current progress. Indian J
Endocrinol Metab 19(6): 731-743.
Sir-Petermann T, Codner E, Maliqueo M, Echiburu B, Hitschfeld C, Crisosto N, PerezBravo F, Recabarren SE & Cassorla F (2006) Increased anti-Mullerian hormone serum
concentrations in prepubertal daughters of women with polycystic ovary syndrome. J
Clin Endocrinol Metab 91(8): 3105-3109.
Skalba P, Cygal A, Madej P, Dabkowska-Huc A, Sikora J, Martirosian G, Romanik M &
Olszanecka-Glinianowicz M (2011) Is the plasma anti-Mullerian hormone (AMH) level
associated with body weight and metabolic, and hormonal disturbances in women with
and without polycystic ovary syndrome? Eur J Obstet Gynecol Reprod Biol 158(2):
254-259.
Smith U (1994) Carbohydrates, fat, and insulin action. Am J Clin Nutr 59(3 Suppl): 686S689S.
Solomon CG, Hu FB, Dunaif A, Rich-Edwards JE, Stampfer MJ, Willett WC, Speizer FE &
Manson JE (2002) Menstrual cycle irregularity and risk for future cardiovascular
disease. J Clin Endocrinol Metab 87(5): 2013-2017.
Sopher AB, Grigoriev G, Laura D, Cameo T, Lerner JP, Chang RJ, McMahon DJ &
Oberfield SE (2014) Anti-Mullerian hormone may be a useful adjunct in the diagnosis
of polycystic ovary syndrome in nonobese adolescents. J Pediatr Endocrinol Metab
27(11-12): 1175-1179.
Spritzer PM, Lecke SB, Satler F, Morsch DM. (2015) Adipose tissue dysfunction,
adipokines, and low-grade chronic inflammation in polycystic ovary syndrome.
Reproduction. 2015 May;149(5):R219-27.
Staessen JA, Fagard R, Thijs L, Celis H, Arabidze GG, Birkenhager WH, Bulpitt CJ, de
Leeuw PW, Dollery CT, Fletcher AE, Forette F, Leonetti G, Nachev C, O'Brien ET,
Rosenfeld J, Rodicio JL, Tuomilehto J & Zanchetti A (1997) Randomised double-blind
comparison of placebo and active treatment for older patients with isolated systolic
hypertension. The Systolic Hypertension in Europe (Syst-Eur) Trial Investigators.
Lancet 350(9080): 757-764.
Stampfer MJ, Krauss RM, Ma J, Blanche PJ, Holl LG, Sacks FM & Hennekens CH (1996)
A prospective study of triglyceride level, low-density lipoprotein particle diameter, and
risk of myocardial infarction. JAMA 276(11): 882-888.
Stanczyk FZ (2006) Diagnosis of hyperandrogenism: biochemical criteria. Best Pract Res
Clin Endocrinol Metab 20(2): 177-191.
Stener-Victorin E, Holm G, Labrie F, Nilsson L, Janson PO & Ohlsson C (2010) Are there
any sensitive and specific sex steroid markers for polycystic ovary syndrome? J Clin
Endocrinol Metab 95(2): 810-819.
103
Stubbs SA, Hardy K, Da Silva-Buttkus P, Stark J, Webber LJ, Flanagan AM, Themmen AP,
Visser JA, Groome NP & Franks S (2005) Anti-mullerian hormone protein expression
is reduced during the initial stages of follicle development in human polycystic ovaries.
J Clin Endocrinol Metab 90(10): 5536-5543.
Sung YA, Oh JY, Chung H & Lee H (2014) Hyperandrogenemia is implicated in both the
metabolic and reproductive morbidities of polycystic ovary syndrome. Fertil Steril
101(3): 840-845.
Taponen S, Martikainen H, Jarvelin MR, Laitinen J, Pouta A, Hartikainen AL, Sovio U,
McCarthy MI, Franks S & Ruokonen A (2003) Hormonal profile of women with selfreported symptoms of oligomenorrhea and/or hirsutism: Northern Finland birth cohort
1966 study. J Clin Endocrinol Metab 88(1): 141-147.
Taponen S, Martikainen H, Jarvelin MR, Sovio U, Laitinen J, Pouta A, Hartikainen AL,
McCarthy MI, Franks S, Paldanius M, Ruokonen A & Northern Finland Birth Cohort
1966 Study (2004) Metabolic cardiovascular disease risk factors in women with selfreported symptoms of oligomenorrhea and/or hirsutism: Northern Finland Birth Cohort
1966 Study. J Clin Endocrinol Metab 89(5): 2114-2118.
Thaler MA, Seifert-Klauss V & Luppa PB (2015) The biomarker sex hormone-binding
globulin - from established applications to emerging trends in clinical medicine. Best
Pract Res Clin Endocrinol Metab 29(5): 749-760.
Tock L, Carneiro G, Pereira AZ, Tufik S & Zanella MT (2014) Adrenocortical production
is associated with higher levels of luteinizing hormone in nonobese women with
polycystic ovary syndrome. Int J Endocrinol 2014: 620605.
Tzeng CR, Chang YC, Chang YC, Wang CW, Chen CH & Hsu MI (2014) Cluster analysis
of cardiovascular and metabolic risk factors in women of reproductive age. Fertil Steril
101(5): 1404-1410.
Ulmer H, Kollerits B, Kelleher C, Diem G & Concin H (2005) Predictive accuracy of the
SCORE risk function for cardiovascular disease in clinical practice: a prospective
evaluation of 44 649 Austrian men and women. Eur J Cardiovasc Prev Rehabil 12(5):
433-441.
van Hooff MH, Voorhorst FJ, Kaptein MB, Hirasing RA, Koppenaal C & Schoemaker J
(2000) Insulin, androgen, and gonadotropin concentrations, body mass index, and waist
to hip ratio in the first years after menarche in girls with regular menstrual cycles,
irregular menstrual cycles, or oligomenorrhea. J Clin Endocrinol Metab 85(4): 13941400.
Vanky E, Kjotrod S, Salvesen KA, Romundstad P, Moen MH & Carlsen SM (2004) Clinical,
biochemical and ultrasonographic characteristics of Scandinavian women with PCOS.
Acta Obstet Gynecol Scand 83(5): 482-486.
Vermeulen A, Verdonck L & Kaufman JM (1999) A critical evaluation of simple methods
for the estimation of free testosterone in serum. J Clin Endocrinol Metab 84(10): 36663672.
Vgontzas AN, Legro RS, Bixler EO, Grayev A, Kales A & Chrousos GP (2001) Polycystic
ovary syndrome is associated with obstructive sleep apnea and daytime sleepiness: role
of insulin resistance. J Clin Endocrinol Metab 86(2): 517-520.
104
Vigier B, Picard JY, Tran D, Legeai L & Josso N (1984) Production of anti-Mullerian
hormone: another homology between Sertoli and granulosa cells. Endocrinology 114(4):
1315-1320.
Visser JA & Themmen AP (2005) Anti-Mullerian hormone and folliculogenesis. Mol Cell
Endocrinol 234(1-2): 81-86.
Wallace IR, McKinley MC, Bell PM & Hunter SJ (2013) Sex hormone binding globulin and
insulin resistance. Clin Endocrinol (Oxf) 78(3): 321-329.
Weenen C, Laven JS, Von Bergh AR, Cranfield M, Groome NP, Visser JA, Kramer P,
Fauser BC & Themmen AP (2004) Anti-Mullerian hormone expression pattern in the
human ovary: potential implications for initial and cyclic follicle recruitment. Mol Hum
Reprod 10(2): 77-83.
Welt CK & Carmina E (2013) Clinical review: Lifecycle of polycystic ovary syndrome
(PCOS): from in utero to menopause. J Clin Endocrinol Metab 98(12): 4629-4638.
West S, Lashen H, Bloigu A, Franks S, Puukka K, Ruokonen A, Jarvelin MR, Tapanainen
JS & Morin-Papunen L (2014) Irregular menstruation and hyperandrogenaemia in
adolescence are associated with polycystic ovary syndrome and infertility in later life:
Northern Finland Birth Cohort 1986 study. Hum Reprod 29(10): 2339-2351.
Wiksten-Almstromer M, Hirschberg AL & Hagenfeldt K (2008) Prospective follow-up of
menstrual disorders in adolescence and prognostic factors. Acta Obstet Gynecol Scand
87(11): 1162-1168.
Wild RA, Carmina E, Diamanti-Kandarakis E, Dokras A, Escobar-Morreale HF, Futterweit
W, Lobo R, Norman RJ, Talbott E & Dumesic DA (2010) Assessment of cardiovascular
risk and prevention of cardiovascular disease in women with the polycystic ovary
syndrome: a consensus statement by the Androgen Excess and Polycystic Ovary
Syndrome (AE-PCOS) Society. J Clin Endocrinol Metab 95(5): 2038-2049.
Wild S, Pierpoint T, McKeigue P & Jacobs H (2000) Cardiovascular disease in women with
polycystic ovary syndrome at long-term follow-up: a retrospective cohort study. Clin
Endocrinol (Oxf) 52(5): 595-600.
Winters SJ, Gogineni J, Karegar M, Scoggins C, Wunderlich CA, Baumgartner R &
Ghooray DT (2014) Sex hormone-binding globulin gene expression and insulin
resistance. J Clin Endocrinol Metab 99(12): E2780-8.
Winters SJ, Talbott E, Guzick DS, Zborowski J & McHugh KP (2000) Serum testosterone
levels decrease in middle age in women with the polycystic ovary syndrome. Fertil
Steril 73(4): 724-729.
World Health Organization multicenter study on menstrual and ovulatory patterns in
adolescent girls. II. Longitudinal study of menstrual patterns in the early
postmenarcheal period, duration of bleeding episodes and menstrual cycles. World
Health Organization Task Force on Adolescent Reproductive Health. (1986) J Adolesc
Health Care 7(4): 236-244.
Wu S, Divall S, Nwaopara A, Radovick S, Wondisford F, Ko C & Wolfe A (2014) Obesityinduced infertility and hyperandrogenism are corrected by deletion of the insulin
receptor in the ovarian theca cell. Diabetes 63(4): 1270-1282.
105
Yildirim B, Sabir N & Kaleli B (2003) Relation of intra-abdominal fat distribution to
metabolic disorders in nonobese patients with polycystic ovary syndrome. Fertil Steril
79(6): 1358-1364.
Yildiz BO (2006) Diagnosis of hyperandrogenism: clinical criteria. Best Pract Res Clin
Endocrinol Metab 20(2): 167-176.
Zawadski J & Dunaif A (1992) Diagnostic criteria for polycystic ovary syndrome. In: Dunaif
et al. Polycystic Ovary Syndrome. Boston, Blackwell Scientific Publications. : 377-384.
Zimmet P, Alberti G, Kaufman F, Tajima N, Silink M, Arslanian S, Wong G, Bennett P,
Shaw J, Caprio S & International Diabetes Federation Task Force on Epidemiology and
Prevention of Diabetes (2007) The metabolic syndrome in children and adolescents.
Lancet 369(9579): 2059-2061.
106
Original publications
I
Pinola P, Lashen H, Bloigu A, Puukka K, Ulmanen M, Ruokonen A, Martikainen H,
Pouta A, Franks S, Hartikainen AL, Järvelin MR & Morin-Papunen L (2012) Menstrual
disorders in adolescence: a marker for hyperandrogenaemia and increased metabolic
risks in later life? Finnish general population-based birth cohort study. Hum Reprod
27(11): 3279-3286.
II Pinola P, Morin-Papunen LC, Bloigu A, Puukka K, Ruokonen A, Järvelin MR, Franks
S, Tapanainen JS & Lashen H (2014) Anti-Müllerian hormone: correlation with
testosterone and oligo- or amenorrhoea in female adolescence in a population-based
cohort study. Hum Reprod 29(10): 2317-2325.
III Pinola P, Piltonen TT, Puurunen J, Vanky E, Sundström-Poromaa I, Stener-Victorin E,
Ruokonen A, Puukka K, Tapanainen JS & Morin-Papunen LC (2015) Androgen Profile
Through Life in Women With Polycystic Ovary Syndrome: A Nordic Multicenter
Collaboration Study. J Clin Endocrinol Metab 100(9): 3400-3407.
IV Pinola P, Puukka K, Piltonen TT, Puurunen J, Vanky E, Sundström-Poromaa I, StenerVictorin E, Lindén Hirschberg A, Ravn P, Skovsager Andersen M, Glintborg D, Roar
Mellembakken J, Ruokonen A, Tapanainen JS & Morin-Papunen LC (2016) Both
normo- and hyperandrogenic women with polycystic ovary syndrome exhibit an
adverse metabolic profile from early adulthood to menopause: a Nordic multicenter
collaboration study. (Manuscript).
Reprinted with permission from Oxford University Press (I and II) and from The
Endocrine Society (III).
Original publications are not included in the electronic version of the dissertation.
107
108
ACTA UNIVERSITATIS OULUENSIS
SERIES D MEDICA
1364. Heikkilä, Vesa-Pekka (2016) New techniques and methods for decreasing healthy
tissue dose in prostate cancer radiotherapy, with special reference to rectal
doses
1365. Aro, Jani (2016) Novel load-inducible factors in cardiac hypertrophy
1366. Myllymäki, Mikko (2016) Hypoxia-inducible factor prolyl 4-hydroxylase-2 in
Tibetan high-altitude adaptation, extramedullary erythropoiesis and skeletal
muscle ischemia
1367. Rubino, Antonino S. (2016) Efficacy of the Perceval sutureless aortic valve
bioprosthesis in the treatment of aortic valve stenosis
1368. Krökki, Olga (2016) Multiple sclerosis in Northern Finland : epidemiological
characteristics and comorbidities
1369. Mosorin, Matti-Aleksi (2016) Prognostic impact of preoperative and
postoperative critical conditions on the outcome of coronary artery bypass
surgery
1370. Pelkonen, Sari (2016) Frozen embryo transfer : early pregnancy, perinatal
outcomes, and health of singleton children
1371. Pohjanen, Vesa-Matti (2016) Toll-like receptor 4 and interleukin 6 gene
polymorphisms in Helicobacter pylori related diseases
1372. Hekkala, Anne (2016) Ketoacidosis at diagnosis of type 1 diabetes in children
under 15 years of age
1373. Kujanpää, Tero (2016) Generalized anxiety disorder and health care utilization
1374. Hintsala, Hanna-Riikka (2016) Oxidative stress and cell adhesion in skin cancer
1375. Lehtonen, Ville (2016) Dental and otologic problems in cleft lip and palate
patients from Northern Finland : cleft associated problems
1376. Koivukangas, Jenni (2016) Brain white matter structure, body mass index and
physical activity in individuals at risk for psychosis : The Northern Finland Birth
Cohort 1986 Study
1377. Väyrynen, Sara (2016) Histological and molecular features of serrated colorectal
adenocarcinoma and its precursor lesions
1378. Kujanpää, Kirsi (2016) Mechanisms behind stem cell therapy in acute myocardial
infarction
1379. Törmälä, Reeta-Maria (2016) Human zona pellucida abnormalities – a genetic
approach to the understanding of fertilization failure
Book orders:
Granum: Virtual book store
http://granum.uta.fi/granum/
D 1380
OULU 2016
UNIVERSITY OF OUL U P.O. Box 8000 FI-90014 UNIVERSITY OF OULU FINLA ND
U N I V E R S I TAT I S
O U L U E N S I S
ACTA
A C TA
D 1380
ACTA
U N I V E R S I T AT I S O U L U E N S I S
Pekka Pinola
University Lecturer Santeri Palviainen
Postdoctoral research fellow Sanna Taskila
Professor Olli Vuolteenaho
University Lecturer Veli-Matti Ulvinen
Director Sinikka Eskelinen
Pekka Pinola
Professor Esa Hohtola
HYPERANDROGENISM,
MENSTRUAL
IRREGULARITIES AND
POLYCYSTIC OVARY
SYNDROME
IMPACT ON FEMALE REPRODUCTIVE AND
METABOLIC HEALTH FROM EARLY ADULTHOOD
UNTIL MENOPAUSE
Professor Jari Juga
University Lecturer Anu Soikkeli
Professor Olli Vuolteenaho
Publications Editor Kirsti Nurkkala
ISBN 978-952-62-1309-5 (Paperback)
ISBN 978-952-62-1310-1 (PDF)
ISSN 0355-3221 (Print)
ISSN 1796-2234 (Online)
UNIVERSITY OF OULU GRADUATE SCHOOL;
UNIVERSITY OF OULU, FACULTY OF MEDICINE;
MEDICAL RESEARCH CENTER OULU;
OULU UNIVERSITY HOSPITAL;
NATIONAL INSTITUTE FOR HEALTH AND WELFARE
D
MEDICA