1 s2.0 S0026049517302743 Main

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 8 6 ( 20 1 8 ) 3 3– 4 3

Available online at www.sciencedirect.com

Metabolism
www.metabolismjournal.com

Female Reproductive Endocrinology

Polycystic ovarian syndrome (PCOS): Long-term


metabolic consequences

Panagiotis Anagnostis a,⁎, Basil C. Tarlatzis a , Robert P. Kauffman b


a
First Department of Obstetrics and Gynecology, Medical School, Aristotle University of Thessaloniki, Thessaloniki, Greece
b
Department of Obstetrics and Gynecology, Texas Tech University Health Sciences Center School of Medicine, Amarillo, TX, USA

A R T I C LE I N FO AB S T R A C T

Article history: Polycystic ovary syndrome (PCOS) is the most common endocrine disorder in women
Received 16 July 2017 during their reproductive ages, associated with a plethora of cardiometabolic
Accepted 26 September 2017 consequences, with obesity, insulin resistance and hyperandrogenemia playing a major
role in the degree of such manifestations. These consequences include increased risk of
Keywords: glucose intolerance and diabetes mellitus (both type 2 and gestational), atherogenic
PCOS dyslipidemia, systemic inflammation, non-alcoholic fatty liver disease, hypertension and
Insulin resistance coagulation disorders. Whether this cluster of metabolic abnormalities is also translated in
Diabetes increased cardiovascular disease (CVD) morbidity and mortality in later life, remains to be
Dyslipidemia established. Data so far based on markers of subclinical atherosclerosis as well as
Fatty liver disease retrospective and prospective cohort studies indicate a possible increased CVD risk, mainly
for coronary heart disease. Future studies are needed to further elucidate this issue.
© 2017 Elsevier Inc. All rights reserved.

1. Introduction have been applied in various studies and at different times.


The first diagnostic attempt was made by the National
Polycystic ovarian syndrome (PCOS) is the most common Institutes of Health (NIH) in 1990 and included both
endocrine disorder in women during their reproductive hyperandrogenism (either clinical or biochemical) and oligo-
period, with a prevalence of 6–20%, depending on the criteria or anovulation, manifested either as frequent (at intervals
used [1,2]. It is the most common cause of menstrual < 21 days) or infrequent bleeding (at intervals > 35 days) [3].
disorders and hirsutism [1,2]. Different diagnostic criteria The Rotterdam criteria were published in 2003 and are the

Abbreviations: AHA, American Heart Association; AEPCOS, Androgen Excess PCOS Society; apoB, apolipoprotein B; apoA-I,
apolipoprotein A-I; CAD, coronary artery disease; CHD, coronary heart disease; CI, confidence interval; CIMT, carotid artery intima-
media thickness; CRP, C-reactive protein; ESE, European Society of Endocrinology; FMD, flow-mediated dilation; FNPO, follicle number
per ovary; GDM, gestational diabetes mellitus; GWAS, genome-wide association studies; HR, hazards ratio; ICAM, intercellular adhesion
molecule; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; IL-6, interleukin-6; IL-18, interleukin-18; IR, insulin resistance;
JIC, Joint Interim Societies; MAPK-ERK, mitogen-activated protein kinase-extracellular signal-regulated kinases; MCP-1, monocyte
chemoattractant protein; MetS, metabolic syndrome; OC, oral contraceptives (OC); OGTT, oral glucose tolerance test; OR, odds ratio;
NAFLD, non-alcoholic fatty liver disease; NCEP ATP III, National Cholesterol Education Program Adult Treatment Panel III; NHLBI,
National Heart Lung and Blood Institute; PAI-1, plasminogen activator inhibitor 1; PCOM, polycystic ovarian morphology; RR, Relative
Risk; T2DM, type 2 diabetes mellitus; TNF-α, tumor necrosis factor-α; WHtR, waist-to-height ratio.
⁎ Corresponding author at: Unit of Reproductive Endocrinology, First Department of Obstetrics and Gynecology, Medical School, Aristotle
University of Thessaloniki, Sarantaporou 10, 54640 Thessaloniki, Greece.
E-mail address: anagnwst@med.auth.gr (P. Anagnostis).

https://doi.org/10.1016/j.metabol.2017.09.016
0026-0495/© 2017 Elsevier Inc. All rights reserved.
34 M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 86 ( 20 1 8 ) 3 3– 43

most commonly adopted. According to these, two of the three hyperandrogenemia or anovulation, PCOS is also associated
following should be fulfilled for diagnosing PCOS: with long-term consequences, such as subfertility, pregnancy
hyperandrogenism (either clinical or biochemical), oligo- or complications, metabolic derangements, such as dysregulation
anovulation and polycystic ovarian morphology (PCOM) of glucose homeostasis, dyslipidemia, coagulation disorders,
(defined either by the presence of ≥ 12 follicles 2–9 mm in fatty liver disease, endothelial dysfunction and potential
maximum diameter or ovarian volume > 10 ml (without a cyst subclinical atherosclerotic disease.
or dominant follicle) in either ovary [1–3]. According to these The purpose of this narrative review was to provide current
criteria four different phenotypes (combinations) may be knowledge regarding the metabolic consequences in patients
detected. It must be noted that the PCOM criterion set by the with PCOS and concomitant cardiovascular disease (CVD) risk.
Rotterdam group may be met by 32% of normal ovulatory
women [4]. Therefore, a higher threshold of follicle number
per ovary (FNPO) of ≥ 25 has been recommended, by using 2. Metabolic Consequences in PCOS
newer technology ultrasound equipment, allowing for maxi-
mal resolution of ovarian follicles. Otherwise, ovarian volume 2.1. Insulin Resistance and Diabetes Mellitus
is suggested instead of FNPO for the diagnosis of PCOM [5].
On the contrary, the Androgen Excess PCOS Society Insulin resistance (IR) is a key player in the metabolic
(AEPCOS) requires the existence of clinical and/or biochemical manifestations in PCOS patients and seems to be in part
hyperandrogenism as a sine qua non for PCOS diagnosis, with independent of obesity. Its prevalence is estimated at 30% of
either polycystic morphology or clinical anovulation as the lean and 70% of obese women with PCOS, acknowledging that
second criterion [6]. The estimated prevalence of PCOS varies “insulin resistance” has been variably defined [12]. Women
according to the criteria used. This was reported to be 8.7% with PCOS have a higher risk of IR and glucose intolerance
using the NIH criteria, rising to 17.8% under the Rotterdam compared with age- and weight-matched women without
criteria and 12% using the AEPCOS definition [7]. No matter PCOS [12,13]. Ethnicity seems also to modify the metabolic
the criteria used, it must be emphasized that the diagnosis of phenotype related to IR. Hispanic PCOS women have a more
PCOS is set after exclusion of other causes of androgen excess, severe phenotype, with a higher prevalence of IR, hypergly-
such as non-classical congenital adrenal hyperplasia, cemia and hyperandrogenism than non-Hispanic white PCOS
hyperprolactinemia, androgen-secreting ovarian or adrenal women. Non-Hispanic black PCOS women have the mildest
tumors, hypothyrodisim, Cushing's syndrome and acromeg- phenotype [14]. A higher body mass index (BMI) exacerbates
aly [1–2]. The diagnosis of PCOS in adolescents in not so IR to a greater extent in women with than those without
straightforward, since anovulation and PCOM are not suffi- PCOS, irrespective of the definition used [13].
cient criteria for this purpose. Thus, clinical and biochemical High testosterone and low sex hormone-binding globulin
evidence of hyperandrogenism, excluding the aforemen- (SHBG) concentrations are independently associated with IR
tioned causes, and persistent oligomenorrhea are required, [13,15]. With respect to the association of IR and PCOS
especially if the latter persists beyond two years after phenotype, it has been shown that the hyperandrogenic and
menarche [2]. No diagnostic criteria exist in peri- and anovulatory phenotype is the most insulin resistant, irrespec-
postmenopausal women. In these cases, a PCOS diagnosis tive of BMI or central adiposity [15]. FNPO is also related to both
can be presumed in the presence of a well-documented IR and biochemical hyperandrogenaemia [15]. Indeed, a large
history of oligomenorrhea and hyperandrogenism during study (n = 1212) in women with PCOS, showed that phenotype
their reproductive period [2]. PCOM is uncommon in post- 1 (all three Rotterdam criteria), is associated with more severe IR
menopausal women given the apoptosis-driven diminution and hyperandrogenemia compared with the other three
in antral follicles [2]. phenotypes. Phenotypes 4 (oligo- or anovulation and PCOM) in
Although the etiology of PCOS has not been fully elucidated, obese/overweight women and 2 (hyperandrogenemia and oligo-
a series of harmful events during fetal and peripubertal life, or anovulation) in both normal and obese/overweight women
which ultimately affect the woman's metabolic and hormonal show greater IR than phenotype 3 (hyperandrogenemia and
phenotype, coupled with genetic defects and environmental PCOM), which does not differ from BMI-matched healthy women
factors, such as obesity and endocrine disrupting chemicals, [16]. Adrenal hyperandrogenemia (DHEAS, Δ4-androstenedione)
appear to play a major role [1,8]. In particular, intrauterine does not seem to deteriorate IR [17].
growth restriction or exposure to androgens has been associ- Many mechanisms have been proposed for the development
ated with the development of PCOS in adult life [9,10], although of IR in PCOS. The main one is a post-binding defect in insulin
others have not found any association between birth weight signaling due to increased serine phosphorylation and de-
and PCOS development [10]. However, no specific interventions creased tyrosine phosphorylation of insulin receptor and
are recommended in PCOS women to prevent the occurrence of insulin receptor substrate-1, which affects metabolic pathways
PCOS in their offspring [2]. Regarding the genetic component of both in classic insulin targets (adipocytes, skeletal muscles) and
PCOS etiology, various genome-wide association studies ovaries. Decreased insulin receptor-β abundance in omental
(GWAS) have been conducted with inconclusive results. These adipose tissue, decreased glucose transporter 4 (GLUT4) in
GWAS involved genes implicated in biosynthesis and activity of subcutaneous adipocytes (leading to decreased glucose uptake),
androgens, metabolism (such as the insulin and insulin- mitochondrial dysfunction, constitutive activation of serine
receptor genes) and inflammatory cytokines, such as tumor kinases in the mitogen-activated protein kinase/extracellular
necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6) genes [11]. signal-regulated kinases (MAPK-ERK) pathway and genetic
Except for the clinical symptomatology associated with disruption of insulin signaling in the central nervous system,
M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 8 6 ( 20 1 8 ) 3 3– 4 3 35

are additional contributors in the development of IR in PCOS 2.2. Dyslipidemia


women. A defect in glucose-stimulated insulin secretion in
PCOS, independent of obesity leading to increased risk of DM A higher prevalence of lipid abnormalities has been reported
has also been reported in PCOS women. In general, in PCOS women with a prevalence of 70% [26]. In general,
hyperinsulinemia is a result of both increased basal insulin dyslipidemia in PCOS is characterized by increased levels of
secretion and decreased hepatic insulin clearance, which low density lipoprotein cholesterol (LDL-c) and very-low
augments androgen production, since insulin acts as a co- density lipoprotein cholesterol (VLDL-c), with high serum
gonadotropin and modulates ovarian steroidogenesis through triglyceride (Tg) and free fatty acid concentrations, as well as
its cognate receptor. Vice versa, androgens alter insulin action decreased high density lipoprotein cholesterol (HDL-c) levels,
(although to a modest extent) in the classic target tissues and particularly HDL2-c, due to decreased apolipoprotein A-I
increase visceral adiposity in PCOS women [18]. (apoA-I) [26–28]. Notably, LDL-c concentrations can be normal
The prevalence of impaired glucose tolerance (IGT) and or borderline high, but this does not correspond to the quality
diabetes mellitus type 2 (T2DM) in PCOS patients is 23–35% and of LDL particles, since LDL exists as subfractions of various
4–10%, respectively, depending also on certain risk factors, such size and atherogenic potential. In PCOS, smaller and dense
as obesity and family history; in lean PCOS the respective LDL particles predominate constituting a more atherogenic
prevalence is 10–15% and 1–2% [18,19]. One of the very few lipid profile converging that of T2DM [29]. PCOS women have
prospective studies on the prevalence of T2DM in European also higher concentrations of oxidized LDL-c, irrespective of
PCOS population showed a cumulative incidence of T2DM of BMI, which further increases CVD risk. Predictors of oxidized
16.9% [incidence rate: 1.05 per 100 person-years (95% CI: LDL-c are apoB/apoA-I ratio in normal weight and TC/HDL-c
0.75–1.41)]. The initial prevalence of T2DM was 2.2% and it ratio in obese PCOS patients. [30]. This atherogenic profile is
rose to 39.3% after a mean follow-up of 19 years. Independent exacerbated by obesity and IR [26–29]. In particular, IR
predictors of the development of diabetes were BMI, fasting increases the hepatic secretion of VLDL (the major source of
glucose and glucose area under the curve at baseline, whereas Tg), decreases the elimination of VLDL and chilomicrons from
the likelihood decreased as SHBG increased [20]. A meta- the circulation and increases the clearance of apolipoprotein
analysis has shown that the odds ratio (OR) of IGT, T2DM and A, the major component of HDL-c [31].
metabolic syndrome (MetS) in PCOS is 2.48 [95% Confidence A meta-analysis showed that, compared with age-
Interval (CI) 1.63, 3.77], 4.43 (95% CI 4.06, 4.82) and 2.88 (95% CI matched controls, PCOS women had 26.39 (95% CI 17.24,
2.40, 3.45), respectively [21]. Subgroup analysis of studies with 35.54) mg/dl higher Tg levels, 6.41 (95% CI 3.69, 9.14) mg/dl
BMI-matched populations reveals the same risks. The OR for lower HDL-c and 18.82 (95% CI 15.53, 22.11) mg/dl higher non-
IGT in lean PCOS women is OR 3.22 (95% CI 1.26, 8.24) [21]. The HDL-c levels. LDL-c levels were also higher even in studies
lifelong risk in a woman with PCOS for developing T2DM with BMI matching [8.32 mg/dl, 95% CI (5.82,10.81)]. The latter
increases with BMI, fasting glucose and glucose area under the remained also higher in normal weight PCOS [9.20 mg/dl,
curve at the initial diagnosis, whereas it decreases as SHBG 95%CI (4.68, 13.72)] [32]. Ethnic and environmental factors
levels increase with age [13,22]. PCOS is also associated with a seem to contribute to differences in the prevalence of lipid
higher risk of developing gestational diabetes mellitus (GDM) disorders amongst PCOS women. In fact, the prevalence of
[in a recent meta-analysis the Relative Risk (RR) was 2.78; 95% low HDL-c (< 50 mg/dl) and high Tg levels in PCOS women
CI: 2.27–3.40; P < 0.001] [23]. Of note, dysregulation of glucose from the USA is estimated at 66.7% and 34.9%, respectively,
homeostasis may be evident even in adolescents with PCOS, whereas the respective percentages in Italian PCOS women
although data are not robust, taking also into account the are 49.1% and 7.4% Tg [33].
discrepancies in the definition used in these women. In one Lipoprotein (a) [Lp(a)] concentrations and apoB (the major
study, the reported prevalence of impaired fasting glucose (IFG), apolipoprotein of atherogenic lipoproteins and indicator of
IGT and T2DM in this subpopulation of PCOS women was 3%, small and dense LDL particles), which are considered indepen-
15.2% and 1.5%, respectively. The frequency of IGT was equal dent risk factors for CVD, may also be elevated in PCOS women
between obese and non-obese PCOS adolescents [24]. [28,34] (in 24% and 14% respectively) [34]. Notably, more than
According to the recent position statement by the European one third of PCOS women with normal plasma lipid profile may
Society of Endocrinology (ESE), an oral glucose tolerance test have elevated Lp(a), apoB or small, dense LDL levels [34].
(OGTT) is recommended in all obese PCOS women and in lean Atherogenic dyslipidemia is more profound in obese
PCOS older than 40 years, with a positive history of GDM or PCOS. Although obesity seems to further worsen lipid profile,
family history of T2DM [1]. Assessment of serum insulin this is regardless of BMI [32]. Some studies have linked
concentrations and insulin resistance indices are not routinely hyperandrogenemia to dyslipidemia [35], but, in others,
recommended [1]. On the contrary, the Endocrine Society [2] testosterone was not associated with PCOS atherogenic
and ESHRE/ASRM [25] recommend the use of OGTT in all lipid profile [29].
adolescents and adult women with PCOS, due to their high risk
in developing IGT and T2DM [2]. Both societies recommend 2.3. Abdominal Obesity-Metabolic Syndrome (MetS)
against the use of hemoglobinA1c (HbA1c), as a screening test
for such purposes, except for the cases of inability or unwill- As shown above, obesity is a key contributor to the clinical
ingness of the patient to complete an OGTT [2]. There is no and metabolic manifestations of PCOS patients. However,
evidence on the optimal time of rescreening, but it has been what is of utmost importance is the body's distribution of
arbitrarily suggested every three to five years, except for an adipose tissue. It is not the increased BMI that matters as
earlier worsening in clinical symptomatology [2]. much as the central fat distribution. Indeed, abdominal
36 M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 86 ( 20 1 8 ) 3 3– 43

obesity has been considered as an independent risk factor for population using the NCEP ATP III criteria the prevalence in
CVD [36]. Comparisons between BMI (an index of general PCOS was 46.4% versus 22.8% in the general population [46].
obesity) and waist-to-height ratio (WHtR) (an index of However, others have reported similar rates with the general
abdominal obesity), has shown that WHtR is more strongly (e.g. Mediterranean) population), that is, 21.4%, applying the
associated with T2DM and metabolic syndrome (MetS). WHtR NCEP ATP III criteria [48]. Interestingly, a large study from the
is also superior to BMI in predicting incidents of CVD, CVD Mediterranean (Greek) population showed higher prevalence
mortality and all-cause mortality [37]. of MetS in PCOS women than in controls only when the NCEP
The prevalence of overweight (BMI: 25–29.9 kg/m2) and ATP III criteria were applied (15.8% versus 10.1%, respectively)
obesity (BMI > 30 kg/m2) in PCOS can reach up to 80%, with a but not with the other MetS definitions (JIC, IDF, AHA/NHLBI)
higher BMI and waist-hip ratiocompared with their age- [49]. The prevalence of MetS also differed according to the
matched controls in the general population [38]. A meta- definition of PCOS. In particular, the prevalence was higher in
analysis showed that PCOS women are at two and three-fold PCOS women defined both by the Rotterdam and NIH criteria
higher risk of being overweight or obese, respectively, compared with controls (regardless of the applied MetS
compared with their non-PCOS counterparts [RR for over- definition), but it was equal when only Rotterdam criteria
weight: 1.95 (95% CI 1.52, 2.50); for obesity: 2.77 (95% CI 1.88, were used [49]. Of note, the former study from the Mediter-
4.10); for central obesity: 1.73 (95% CI: 1.31, 2.30)]. This ranean (Spanish) population used the NIH criteria [48].
prevalence is also affected by ethnicity, since it is higher in Therefore, no safe conclusions can be made with respect to
Caucasian compared with Asian women [RR: 10.79 (5.36, the prevalence of MetS in PCOS due to the heterogeneity of
21.70) versus 2.31 (1.33, 4.00), respectively] [39]. Adipose fat definitions used and the populations studied.
distribution in PCOS shifts to android pattern, both in lean Contributors to the pathogenesis of obesity in PCOS are some
and obese women, compared with non-PCOS women [38–41]. mutations in the peroxisome proliferator activated receptor-γ
Moreover, lean PCOS women are characterized by a lower (PPAR-γ) gene [50], which enhance adipogenesis, eating behavior
amount of body fat in the upper region and a higher leg (mainly due to the association of hyperandorgenism with
subcutaneous adipose tissue reduction, resembling the fat increased appetite) [51], reduced secretion of cholecystokinin,
pattern in children before their sex specific body fat distribu- which reduces satiety [52], and increased levels of neuropeptide
tion. On the other hand, obese PCOS women exhibit a Y, which increases appetite (perhaps due to dysregulated
significantly increased upper trunk obesity, appearing as secretion of ghrelin both in obese and lean PCOS women) [53].
“super-apples”, resembling the T2DM pattern [40,41]. Obesity Androgens seem to play a significant role in the pathogenesis of
and concomitant hyperinsulinemia further increases bio- MetS in PCOS. The prevalence of MetS is higher in
available androgen levels, mainly due to decreased concen- hyperandrogenic than non-hyperandrogenic anovulatory
trations of SHBG [38]. On the other hand, the prevalence of women with PCOS (24.8% vs 0%). Interestingly, free androgen
PCOS is higher (about five-fold) in premenopausal overweight index and waist circumference are independently associated
and obese women compared with normal-weight from the with the presence of MetS and/or IR [54]. Accordingly, in another
general population [42]. Hyperandrogenemia may also con- study, the non-hyperandrogenic PCOS phenotype was associated
tribute to abdominal adiposity in PCOS women [38]. with a milder phenotype compared with the other phenotypes
“MetS” is a term used to incorporate a cluster of metabolic according to the Rotterdam criteria [the respective prevalences of
abnormalities, including abdominal obesity, hyperglycemia, MetS in the full PCOS (three criteria), oligo-and/or anovulatory
hypertension and dyslipidemia, with IR playing a pivotal role hypeandrogenic, ovulatory hyperandrogenic PCOM and ovulato-
in its pathogenesis. MetS is associated with a five-fold higher ry PCOM non-hyperandrogenic phenotypes were 28.5%, 25.5%,
risk for T2DM and two-fold higher risk for CVD [43]. It 8.3% and 7.2%] [55].
generally affects about 20–25% of the Mediterranean popula- Visceral obesity is associated with increased levels of
tion [43]. The prevalence of MetS depends on the definition proinflammatory factors such as TNF-α, IL-6 and
used. Different criteria have been proposed by various interleukin-18 (IL-18), which are secreted by activated tissue
organizations. The latest were proposed by the Joint Interim marcophages that infiltrate adipose tissue and seem to be
Societies (JIC) and require at least three of the following: implicated in the pathogenesis of IR [56]. Increased concen-
central obesity (waist circumference > 94 cm for men and trations of TNF-α [57] have been detected in PCOS women,
> 80 cm for women in the Mediterranean population); Tg irrespective of obesity, as well as IL-6 [58] andIL-18 [59],
levels > 1.7 mmol/l (150 mg/dl); HDL-c levels > 1.0 mmol/l although not by all studies [60]. Moreover, IL-6 gene polymor-
(40 mg/dl) for men and > 1.3 mmol/l (50 mg/dl) for women; phisms have been associated with total testosterone and lipid
fasting blood glucose levels > 6.1 mmol/l (110 mg/dl) [44]. With parameters (such as HDL-c and Tg) in PCOS women [61].
the JIC criteria, the prevalence of MetS rises up to 45.7%, but is Increased systemic inflammation, as indicated by increased
less predictive of CVD compared with that based on the concentrations of C-reactive protein (CRP), has been demon-
International Diabetes Federation (IDF), National Cholesterol strated in PCOS, compared with BMI-matched controls [62].
Education Program Adult Treatment Panel III (NCEP ATP III) CRP, more specifically high-sensitivity CRP, has been recog-
and the American Heart Association/National Heart Lung and nized as an independent predictor of CVD [63]. Whether this
Blood Institute (AHA/NHLBI) definition (respective prevalence systemic inflammation plays a causal role in the pathogen-
43.5%, 24.5% and 26.3%) [44]. The prevalence of MetS is higher esis of PCOS or is just a consequence of the metabolic and
in PCOS compared with the general population (ranging from hormonal milieu remains to be clarified. In general, there is
33% to 47% in most studies, increasing to 53% in ages 30– an adipose tissue functional derangement in PCOS patients.
39 years) [45–47]. In a representative study from the US Adipose hypertrophy has been reported by several studies,
M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 8 6 ( 20 1 8 ) 3 3– 4 3 37

associated with hyperandrogenemia, which further contrib- 2.5. Other Metabolic Consequences in PCOS
utes to the pathogenesis of IR. Adipocyte hypertrophy is also
associated with reduced catecholamine induced lipolysis in 2.5.1. Non-alcoholic Fatty Liver Disease (NAFLD)
subcutaneous lipolysis adipocytes [64]. NAFLD constitutes a wide spectrum of pathologic conditions
from simple steatosis, characterized by lipid accumulation in
2.4. Adipokines the liver parenchyma to non-alcoholic steatohepatitis
(NASH), characterized by chronic inflammation and fibrosis
A link between adiposity, IR, systemic inflammation and [73]. The prevalence of NAFLD has increasingly risen during
concomitant increased risk of atherosclerosis is provided by a the last decades, affecting 6–45% of the general population. It
group of cytokines, secreted by adipose tissue, also called is even higher in patients with T2DM (70%) and in morbidly
“adipokines”. The main representatives of this category are obese patients (up to 90%). NAFLD and NASH have also been
adiponectin and leptin. The former exerts insulin sensitizing associated with increased risk of CVD. NAFLD can rarely lead
and anti-inflammatory properties; low adiponectin circulat- to cirrhosis and hepatocellular carcinoma. Despite the recent
ing levels are usually found in obesity and MetS [65]. A meta- progress in indices and techniques diagnosing different
analysis has shown lower adiponectin concentrations in both stages of NAFLD, liver biopsy remains the gold standard
lean and obese PCOS women compared with their healthy method [73].
counterparts, regardless of the degree of obesity [66]. Specific Increased prevalence of NAFLD has also been reported in
adiponectin gene polymorphisms are associated with higher PCOS women (27–62%) [74–76]. Notably, an increased preva-
(T45G) or lower (G276T) risk of PCOS, highlighting a specific lence of PCOS has been also described in females of reproduc-
role of this adipokine in the pathogenesis or metabolic tive age with NAFLD [75], suggesting an interplay between these
consequences of PCOS [67]. two conditions or common pathogenetic pathways. IR seems to
In contrast to adiponectin, leptin has a central and peripheral play the main pathogenetic role as it has been associated with
regulatory role, by increasing satiety and controlling appetite and the development of NAFLD independently of obesity (BMI or
energy expenditure, according to the levels of energy stored as waist circumference) [76]. Serum androgens seem also to
body fat. High leptin concentrations are associated with systemic independently predict NAFLD irrespective of other con-
inflammation, IR and high risk of atherosclerosis [65]. A meta- founders, such as BMI. A recent meta-analysis reported an OR
analysis showed higher leptin levels in PCOS women compared of 2.54 (95% CI 2.19–2.95), which is higher in those with
with their healthy counterparts. This difference was evident in hyperandrogenism (classic phenotype) compared to PCOS
obese PCOS only, compared with obese non-PCOS, regardless of women without hyperandrogenism [77]. Despite the well-
BMI. In the non-obese PCOS, leptin levels were moderately higher established interplay between obesity, IR and NAFLD, this and
compared with non-obese controls [68]. Leptin may also contrib- other studies [78] highlight the significant and unique contri-
ute to hyperandrogenism by promoting steroidogenesis and bution of hyperandrogenemia to the metabolic phenotype of
inhibiting neuropeptide Y, which leads to high gonadotropin PCOS women. Of course, the presence of IR per se leading to
releasing hormone (GnRH) and lutenizing hormone (LH) levels hyperandrogenemia further contributes indirectly to the path-
[69]. Other adipokines implicated in the pathogenesis of insulin ogenesis of NAFLD in PCOS patients. Notably, the prevalence of
resistance are resistin, visfatin, apelin, vaspin, chenmerin and advanced liver disease (NASH with fibrosis) in women with
omentin-1 [65], although conflicting data exist regarding their PCOS is higher than their non-PCOS counterparts [79].
role in PCOS. Regarding resistin, differences between PCOS and
non-PCOS are not robust, except perhaps in some populations 2.5.2. Hypertension
(i.e. Chinese) [64]. Interestingly, a local role of resistin in PCOS Although data are not robust, PCOS women seem to be at
pathogenesis cannot be excluded, since increased gene expres- increased risk for hypertension, at least in later post-
sion has been detected in adipocytes of PCOS patients [64]. reproductive life. In general, the prevalence of hypertension
Regarding visfatin, contradictory findings exist in studies in PCOS (mainly systolic) in PCOS premenopausal women is estimated
populations, despite its association with IR, pro-inflammatory at 9–25.7%, higher than the general population [80–83]. Older
markers and endothelial dysfunction [64,70]. However, increased studies showed higher mean arterial pressure and ambulato-
gene expression has been reported in both omental and ry systolic pressure in PCOS compared with non-PCOS
subcutaneous adipose tissue [64]. Moreover, a recent meta- subjects, independently of BMI and IR [83] or no difference at
analysis showed higher concentrations in PCOS compared with all in these parameters [84]. Alarmingly, the physiological
non-PCOS women, irrespective of BMI, IR and total testosterone nocturnal drop in both systolic and diastolic pressure may not
levels [71]. Regarding omentin-1, lower levels have been found be observed in 51% and 23% of PCOS women, respectively [82].
in PCOS compared with healthy controls, by another Moreover, some studies in postmenopausal women support
recent meta-analysis, associated with HOMA-IR, but without the notion that the presence of PCOS during reproductive ages
correlation with BMI and total testosterone concentrations confers a higher risk for developing hypertension after
[72]. With respect to the other adipokines conflicting data menopause [81,85]. Except for obesity and IR, an independent
exist in PCOS women [64]. pathogenetic role of androgens in the prevalence of hyperten-
An intriguing role in the pathogenesis of PCOS and its sion has been suggested through activation of the renin-
metabolic consequences, such as IR, emerges for adipose tissue angiotensin system [86]. Moreover, others found higher preva-
and its derived adipokines. In addition to their implication in IR lence of hypertension in both normo- and hyperandrogenic
and atherosclerotic disease, these molecules may also affect women with PCOS compared with healthy controls, indepen-
sex steroid production and fertility in PCOS women. dently from BMI, suggesting other pathogenetic pathways [87].
38 M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 86 ( 20 1 8 ) 3 3– 43

2.5.3. Coagulation Disorders parameter cannot improve CVD risk prediction, independently
PCOS has been associated with disturbances in the coagulation of other traditional factors, such as hypertension, diabetes and
and fibrinolysis systems, suggesting a potential prothrombotic dyslipidemia [99].
state, although data are conflicting and emerge from small Another significant parameter is that the prevalence of
studies. Increased levels of plasminogen activator inhibitor 1 CVD in the young population is very low and strongly
(PAI-1) and fibrinogen have been found in PCOS compared with increases during the 50s [1,2]. Several studies indicate an
healthy controls, irrespective of BMI. These disorders of increased subclinical atherosclerosis in PCOS, by various
fibrinolysis and coagulation are associated with low SHBG and indices such as coronary artery calcification scores [60,100],
hyperinsulinemia [88]. Impaired global fibrinolytic capacity has carotid artery intima-media thickness (CIMT) [101,102] and
also been confirmed by others using a method based on the endothelial dysfunction measured by flow-mediated dila-
amount of generated D-dimers when fibrinolysis is stopped by tion (FMD) [101,103], compared with controls, even from
adding aprotinin [89]. However, others did not confirm the early reproductive ages. Arterial stiffness, another marker
difference in PAI-1 [90]. Regarding other coagulation factors, of CVD, is also increased in PCOS women compared with
such as von Willebrand's factor, factor II, factor VII, protein C, their non-PCOS counterparts [104]. Meta-analyses have
protein S, factor V Leiden, prothrombin G20210A and anti- confirmed these results, despite increased heterogeneity
thrombin, no difference between PCOS and non-PCOS has been between studies [105,106]. Interestingly, in one of them,
demonstrated [89,91,92]. Finally, increased homocysteine con- the decreased FMD in PCOS was not affected by age and
centrations have been demonstrated in PCOS, independent of BMI [106].
BMI [93]. It must also be noted that most of these studies were
The effect of oral contraceptives (OC) in augmenting the limited by small sample sizes and included young women
risk of venous thromboembolism should also be taken into of reproductive age. The presence of venous thromboem-
account in women with PCOS (three- to six-fold higher RR, bolism is also associated with a lower risk of developing
although the absolute risk seems to be low, depending on the atherosclerotic CVD [107]. Except for the traditional CVD
estrogen dose and the generation of OC (higher with 30 μg/d risk factors, androgens may be implicated in the devel-
of ethinyl-estradiol and the third OC generation) [94]). opment of CVD, although it is not clear if they play a
Patient's characteristics, such as age, BMI, smoking and protective or detrimental role on the cardiovascular
positive family history for thrombophilia should also be system. Total testosterone and Δ4-androstenedione con-
taken into account. BMI does not seem to be significantly centrations are positively, whereas DHEAS and SHBG are
affected by OC or anti-androgens [94]. OC may also increase IR negatively, associated with CIMT [102,103]. However,
and alter lipid profile, by increasing Tg and HDL-c and others found no association between clinical and/or
decreasing LDL-c/HDL-c ratio, although data are inconsistent biochemical hyperandrogenemia and atherosclerotic dis-
and contradictory, depending again on the dose of OC and ease [104].
patient's characteristics (BMI, age, family history of diabetes) Regarding the difference in the incidence of CVD events
[94–96]. OC may also increase systemic inflammation, as between PCOS and non-PCOS, few studies exist. In a
indicated by the levels of intercellular adhesion molecule retrospective population based cohort study from Australia
(ICAM)-1, TNF-α and monocyte chemoattractant protein (n = 2566), PCOS women (median age 35.8 years) had more
(MCP)-1 [96]. OC may be weakly associated with increased hospitalizations for ischemic heart disease (0.8 versus 0.2%)
CVD risk, although the quality of evidence is low [94]. and cerebrovascular disease (0.6 versus 0.2%) compared with
Progestogen only pills or hormone releasing intrauterine their non-PCOS counterparts [108]. It must be emphasized
devices do not increase the risk of venous thromboembolism. that the CVD risk in young ages is very low and a
The risk is increased only when progestogen is combined with hospitalization study may overestimate this phenomenon.
ethinyl-estradiol and is lowest with levonogestrel [97]. One study that reviewed death certificates of 786 women in
the UK (with a diagnosis of PCOS at an average age of
26.4 years and a mean follow-up time of 30 years) failed to
show a statistically significant increase in cardiovascular
3. Cardiovascular Risk in PCOS. Is it Indeed mortality [109]. However, it must be noted that the
Increased? diagnosis of PCOS was based on historical data during a
period of 49 years (1930–1979) and was not supported by
All the aforementioned data suggest an increased cardio- hormonal assessments. The same authors in a later study,
metabolic risk profile in PCOS women. However, if this risk is including 309 postmenopausal women, did not show a
translated into increased risk for cardiovascular events difference in all-cause and CVD mortality neither in
remains to be established, taking also into account that the coronary heart disease (CHD), but they did notice a higher
phenotype of PCOS may improve with aging. Well designed, OR for cerebrovascular disease in PCOS women [2.8 (95% CI
high quality longitudinal studies are needed, specifically 1.1–7.1)] [110]. Another retrospective cohort study (n = 309,
focusing on the association of the clinical aspects of PCOS, mean age 43.7 years, follow-up time 23.7 years) did not
such as abdominal obesity and IR with CVD risk. Most show any difference in CVD events between PCOS and non-
prospective studies in the general population have PCOS women [111]. However, another retrospective cohort
shown that hyperinsulinemia is independently associat- study from the UK, including 2301 PCOS women (mean
ed with the presence of atherosclerosis and CHD [98]. age 36.3 years) followed for 20 years, showed higher
The evidence is not robust for abdominal obesity. This incidence of myocardial infarction and angina compared
M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 8 6 ( 20 1 8 ) 3 3– 4 3 39

Table 1 – Studies reporting data on cardiovascular disease events in women with PCOS.
Study PCOS (n) Age (yrs) Follow-up (yrs) Design CHD events CVD events Other/total CVD
(author, year) (OR or HR) (OR or HR) events (OR or HR)

Hart, 2015 2566 35.8 14 Retrospective cohort


2.89 (1.68–4.97) 2.58 (1.43–4.67)
Pierpoint, 1998 786 20–74 30 Retrospective cohort
1.40 (0.75–2.40) 0.23 (0.03–0.85)
Wild, 2000 319 56.7 31 Retrospective cohort
1.5 (0.7–2.9) 2.8 (1.1–7.1)
Mani, 2013 2301 36.3 20 Retrospective cohort
From 2.6 (1.0–6.3) to
12.9 (3.4–8.6)
Iftikhar, 2012 309 46.7 23.7 Retrospective cohort 0.74 (0.32–1.72) 1.05 (0.28–3.92) 5.67 (0.51–63.7)
Schmidt, 2011 35 61–79 21 Prospective cohort No difference No difference
with controls with controls
Merz, 2016 25 62.6 9.3 Prospective cohort 0.82 (0.37–1.81)
Wang, 2011 1974 26.9 40 Prospective cohort 1.42 (1.03–1.94) 0.85 (0.49–1.47) 1.21 (0.97–1.52)
[After adjustment for
BMI: 1.35 (0.98–1.85)]
Solomon, 2002 82,479 47.8–48.6 14 Prospective cohort Non-fatal CHD: 1.30 (0.97–1.74)
1.25 (1.07–1.47)
Fatal CHD: 1.67
(1.35–2.06)
de Groot, 2011 83,061 Meta-analysis 2.02 (1.47–2.76)
Zhao, 2016 104,392 20–73.8 7–40 Meta-analysis 1.44 (1.13–1.84) 1.30 (1.09–1.56)

PCOS: polycystic ovary syndrome; CHD: coronary heart disease; CVD: cardiovascular disease; OR: odds ratio; HR: hazard ratio.

with the local female population, with the highest OR in No significant difference was found for CVD and cerebrovas-
the group of >65 years old [from 2.6 (95% CI: 1.0–6.3) to 12.9 (CI: cular mortality [114]. However, in another prospective cohort
3.4–48.6)] [112]. study (n = 82,479, from the Nurses' Health Study) conducted
Data from prospective cohort studies provide stronger on the same concept, women with irregular cycles (n = 1974)
evidence. In one of them, no difference in myocardial did have an increased risk for non-fatal and fatal CHD
infarction, stroke and mortality was reported between 35 mortality (age adjusted RR 1.25 and 1.67, respectively; 95% CI
PCOS women (61–79 years) and 120 age-matched controls, 1.07–1.47 and 1.35–2.06, respectively), but without difference
after 21 years of follow-up [85]. In another (n = 295) cohort for stroke, compared to women with regular menses. This risk
study of postmenopausal women, the diagnosis of PCOS remained significant after adjustment for BMI [115].
(defined by their premenopausal history) was not associated To better clarify whether CVD risk is indeed increased in
with angiographically documented coronary artery disease PCOS compared with non-PCOS women, two meta-analyses
(CAD) or mortality [113]. Conversely, another prospective have been conducted so far. The first included five controlled
cohort study demonstrated that women with irregular cycles observational studies, published between 2000 and 2008. The
(n = 1974) had an increased risk for CHD mortality [age RR for CHD, including fatal and non-fatal myocardial infarction
adjusted hazards ratio (HR) 1.42, 95% CI 1.03–1.94] compared or stroke was 2.02 (95% CI 1.47–2.76), which remained signifi-
with women with regular menses. This risk was attenuated cant after adjustment for BMI [RR: 1.55 (95% CI 1.27–1.89)] [116].
after adjustment for BMI (adjusted HR 1.35, 95% CI 0.98–1.85). The other more recent meta-analysis included five case-control

Fig. 1 – Pathogenesis of cardiovascular disease risk in PCOS women.


40 M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 86 ( 20 1 8 ) 3 3– 43

and five cohort studies. According to this, the OR for CVD in [5] Dewailly D, Lujan ME, Carmina E, Cedars MI, Laven J,
PCOS was 1.30 (95% CI 1.09–1.56), without any difference with Norman RJ, et al. Definition and significance of polycystic
ovarian morphology: a task force report from the Androgen
respect to study design. Regarding the type of CVD, the OR was
Excess and Polycystic Ovary Syndrome Society. Hum Reprod
increased for CHD (1.44; 95% CI 1.13–1.84) [117]. The data from Update 2014;20:334–52.
the aforementioned studies are presented briefly in Table 1. [6] Azziz R, Carmina E, Dewailly D, Diamanti-Kandarakis E,
In general, in aged PCOS women the available data are not Escobar-Morreale HF, Futterweit W, et al. The Androgen
sufficient for safe conclusions. The pathogenesis of the Excess and PCOS Society criteria for the polycystic ovary
potentially increased CVD risk in PCOS are presented in Fig. 1. syndrome: the complete task force report. Fertil Steril 2009;
The main findings suggest that the number of CVD events in 91:456–88.
[7] March WA, Moore VM, Willson KJ, Phillips DI, Norman RJ,
aged PCOS patients is similar to those without PCOS or only
Davies MJ. The prevalence of polycystic ovary syndrome in a
marginally increased, but it is lower than that expected on the community sample assessed under contrasting diagnostic
basis of risk calculation during younger ages. In general, there is criteria. Hum Reprod 2010;25:544–51.
a discrepancy between CVD risk and late events in aged PCOS, [8] Goodarzi MO, Dumesic DA, Chazenbalk G, Azziz R. Polycys-
the mechanisms for which are unclear. Progressive normaliza- tic ovary syndrome: etiology, pathogenesis and diagnosis.
tion of CVD risk during late reproductive age may play a main Nat Rev Endocrinol 2011;7:219–31.
[9] Sir-Petermann T, Hitchsfeld C, Maliqueo M, Codner E,
role. Furthermore, in the majority of cases, the severity of PCOS
Echiburú B, Gazitúa R, et al. Birth weight in offspring of
reduces during lifetime, perhaps due to spontaneous reduction
mothers with polycystic ovarian syndrome. Hum Reprod
of ovarian and adrenal androgens. The emergence of ovulatory 2005;20:2122–6.
cycles with aging leads to improvement in lipid profile and [10] Paschou SA, Ioannidis D, Vassilatou E, Mizamtsidi M,
reduction in prevalence of MetS [1,2]. Panagou M, Lilis D, et al. Birth weight and polycystic ovary
syndrome in adult life: is there a causal link? PLoS One 2015;
10:e0122050.
[11] Zhao H, Lv Y, Li L, Chen ZJ. Genetic studies on polycystic
4. Conclusions
ovary syndrome. Best Pract Res Clin Obstet Gynaecol 2016;
37:56–65.
PCOS is associated with significant metabolic consequences, [12] Randeva HS, Tan BK, Weickert MO, Lois K, Nestler JE, Sattar
mediated by obesity and IR, but also by independent N, et al. Cardiometabolic aspects of the polycystic ovary
pathways. These consequences include increased risk of IGT syndrome. Endocr Rev 2012;33:812–41.
and T2DM, atherogenic dyslipidemia, systemic inflammation [13] Cassar S, Misso ML, Hopkins WG, Shaw CS, Teede HJ, Stepto NK.
due to increased secretion of pro-inflammatory factors by Insulin resistance in polycystic ovary syndrome: a systematic
review and meta-analysis of euglycaemic-hyperinsulinaemic
adipose tissue, NAFLD, hypertension and potential coagula-
clamp studies. Hum Reprod 2016;31:2619–31.
tion disorders. Whether this apparent increased CVD risk in [14] Engmann L, Jin S, Sun F, Legro RS, Polotsky AJ, Hansen KR,
reproductive ages is translated to increased CVD morbidity et al. Reproductive Medicine Network. Racial and ethnic
and mortality in later life, remains to be established. Data so differences in the polycystic ovary syndrome metabolic
far, based on markers of subclinical atherosclerosis, some phenotype. Am J Obstet Gynecol 2017;216:493.e1–493.e13.
retrospective and prospective cohort studies, indicate a [15] Diamanti-Kandarakis E, Panidis D. Unravelling the pheno-
typic map of polycystic ovary syndrome (PCOS): a prospec-
possible increased CVD risk, mainly for CHD.
tive study of 634 women with PCOS. Clin Endocrinol (Oxf)
2007;67:735–42.
[16] Panidis D, Tziomalos K, Misichronis G, Papadakis E, Betsas G,
Katsikis I, et al. Insulin resistance and endocrine characteris-
Conflict of Interest tics of the different phenotypes of polycystic ovary syndrome:
a prospective study. Hum Reprod 2012;27:541–9.
The authors have no conflict of interest to disclose. [17] Paschou SA, Palioura E, Ioannidis D, Anagnostis P,
Panagiotakou A, Loi V, et al. Adrenal hyperandrogenism
does not deteriorate insulin resistance and lipid profile in
women with PCOS. Endocr Connect 2017;6:601–6.
REFERENCES
[18] Diamanti-Kandarakis E, Dunaif A. Insulin resistance and the
polycystic ovary syndrome revisited: an update on mecha-
[1] Conway G, Dewailly D, Diamanti-Kandarakis E, Escobar- nisms and implications. Endocr Rev 2012;33:981–1030.
Morreale HF, Franks S, Gambineri A, et al, ESE PCOS Special [19] Legro RS, Kunselman AR, Dodson WC, Dunaif A. Prevalence
Interest Group. The polycystic ovary syndrome: a position and predictors of risk for type 2 diabetes mellitus and
statement from the European Society of Endocrinology. Eur J impaired glucose tolerance in polycystic ovary syndrome: a
Endocrinol 2014;171:P1–P29. prospective, controlled study in 254 affected women. J Clin
[2] Legro RS, Arslanian SA, Ehrmann DA, Hoeger KM, Murad MH, Endocrinol Metab 1999;84:165–9.
Pasquali R, et al, Endocrine Society. Diagnosis and treatment [20] Gambineri A, Patton L, Altieri P, Pagotto U, Pizzi C, Manzoli L,
of polycystic ovary syndrome: an Endocrine Society clinical et al. Polycystic ovary syndrome is a risk factor for type 2
practice guideline. J Clin Endocrinol Metab 2013;98:4565–92. diabetes: results from a long-term prospective study.
[3] Lujan ME, Chizen DR, Pierson RA. Diagnostic criteria for Diabetes 2012;61:2369–74.
polycystic ovary syndrome: pitfalls and controversies. [21] Moran LJ, Misso ML, Wild RA, Norman RJ. Impaired glucose
J Obstet Gynaecol Can 2008;30:671–9. tolerance, type 2 diabetes and metabolic syndrome in
[4] Johnstone EB, Rosen MP, Neril R, Trevithick D, Sternfeld B, Murphy polycystic ovary syndrome: a systematic review and meta-
R, et al. The polycystic ovary post-rotterdam: a common, age- analysis. Hum Reprod Update 2010;16:347–63.
dependent finding in ovulatory women without metabolic [22] Livadas S. Letter to the Editor: Development and risk factors
significance. J Clin Endocrinol Metab 2010;95:4965–72. of type 2 diabetes in a nation-wide population of women
M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 8 6 ( 20 1 8 ) 3 3– 4 3 41

with polycystic ovary syndrome. J Clin Endocrinol Metab [41] Tafeit E, Möller R, Rackl S, Giuliani A, Urdl W, Freytag U, et al.
2017 Nov 3. https://doi.org/10.1210/jc.2017-02051 [Epub Subcutaneous adipose tissue pattern in lean and obese
ahead of print]. women with polycystic ovary syndrome. Exp Biol Med
[23] Yu HF, Chen HS, Rao DP, Gong J. Association between (Maywood) 2003;228:710–6.
polycystic ovary syndrome and the risk of pregnancy [42] Alvarez-Blasco F, Botella-Carretero JI, San Millan JL, Escobar-
complications: a PRISMA-compliant systematic review and Morreale HF. Prevalence and characteristics of the polycys-
meta-analysis. Medicine (Baltimore) 2016;95:e4863. tic ovary syndrome in overweight and obese women. Arch
[24] Flannery CA, Rackow B, Cong X, Duran E, Selen DJ, Burgert Intern Med 2006;166:2081–6.
TS. Polycystic ovary syndrome in adolescence: impaired [43] Anagnostis P. Metabolic syndrome in the Mediterranean
glucose tolerance occurs across the spectrum of BMI. Pediatr region: current status. Indian J Endocrinol Metab 2012;16:
Diabetes 2013;14:42–9. 72–80.
[25] Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus [44] Athyros VG, Ganotakis ES, Tziomalos K, Papageorgiou AA,
Workshop Group. Revised 2003 consensus on diagnostic Anagnostis P, Griva T, et al. Comparison of four definitions
criteria and long-term health risks related to polycystic of the metabolic syndrome in a Greek (Mediterranean)
ovary syndrome. Fertil Steril 2004;81(1):19–25. population. Curr Med Res Opin 2010;26:713–9.
[26] Legro RS, Kunselman AR, Dunaif A. Prevalence and predic- [45] Essah PA, Nestler JE. The metabolic syndrome in polycystic
tors of dyslipidemia in women with polycystic ovary ovary syndrome. J Endocrinol Invest 2006;29:270–80.
syndrome. Am J Med 2001;111:607–13. [46] Glueck CJ, Papanna R, Wang P, Goldenberg N, Sieve-Smith L.
[27] Wild RA, Painter PC, Coulson PB, Carruth KB, Ranney GB. Incidence and treatment of metabolic syndrome in newly
Lipoprotein lipid concentration and cardiovascular risk in referred women with confirmed polycystic ovarian syn-
women with polycystic ovary syndrome. J Clin Endocrinol drome. Metabolism 2003;52:908–15.
Metab 1985;61:946–51. [47] Vryonidou A, Paschou SA, Muscogiuri G, Orio F, Goulis DG.
[28] Yilmaz M, Biri A, Bukan N, Karakoç A, Sancak B, Törüner F, Mechanisms in endocrinology: metabolic syndrome
et al. Levels of lipoprotein and homocysteine in non-obese through the female life cycle. Eur J Endocrinol 2015;173:
and obese patients with polycystic ovary syndrome. R153–63.
Gynecol Endocrinol 2005;20:258–63. [48] Espinós-Gómez JJ, Rodriguez-Espinosa J, Ordóñez-Llanos J,
[29] Pirwany IR, Fleming R, Greer IA, Packard CJ, Sattar N. Lipids Calaf-Alsina J. Metabolic syndrome in Mediterranean
and lipoprotein subfractions in women with PCOS: rela- women with polycystic ovary syndrome: when and how to
tionship to metabolic and endocrine parameters. Clin predict its onset. Gynecol Endocrinol 2012;28:264–8.
Endocrinol (Oxf) 2001;54:447–53. [49] Panidis D, Macut D, Tziomalos K, Papadakis E, Mikhailidis K,
[30] Macut D, Damjanovic S, Panidis D, Spanos N, Glisic B, Kandaraki EA, et al. Prevalence of metabolic syndrome in
Petakov M, et al. Oxidised low-density lipoprotein concen- women with polycystic ovary syndrome. Clin Endocrinol
tration - early marker of an altered lipid metabolism in (Oxf) 2013;78:586–92.
young women with PCOS. Eur J Endocrinol 2006;155:131–6. [50] Orio Jr F, Matarese G, Di Biase S, Palomba S, Labella D, Sanna
[31] Savage DB, Petersen KF, Shulman GI. Disordered lipid V, et al. Exon 6 and 2 peroxisome proliferator-activated
metabolism and the pathogenesis of insulin resistance. receptor-gamma polymorphisms in polycystic ovary syn-
Physiol Rev 2007;87:507–20. drome. J Clin Endocrinol Metab 2003;88:5887–92.
[32] Wild RA, Rizzo M, Clifton S, Carmina E. Lipid levels in [51] Hirschberg AL, Naessén S, Stridsberg M, Byström B, Holtet J.
polycystic ovary syndrome: systematic review and meta- Impaired cholecystokinin secretion and disturbed appetite
analysis. Fertil Steril 2011;95 [1073-9.e1-11]. regulation in women with polycystic ovary syndrome.
[33] Berneis K, Rizzo M, Hersberger M, Rini GB, Di Fede G, Pepe I, Gynecol Endocrinol 2004;19:79–87.
et al. Atherogenic forms of dyslipidaemia in women with [52] Naessén S, Carlström K, Garoff L, Glant R, Hirschberg AL.
polycystic ovary syndrome. Int J Clin Pract 2009;63:56–62. Polycystic ovary syndrome in bulimic women–an evaluation
[34] Essah PA, Nestler JE, Carmina E. Differences in dyslipidemia based on the new diagnostic criteria. Gynecol Endocrinol
between American and Italian women with polycystic ovary 2006;22:388–94.
syndrome. J Endocrinol Invest 2008;31:35–41. [53] Moran LJ, Noakes M, Clifton PM, Wittert GA, Tomlinson L,
[35] Graf MJ, Richards CF, Brown V, Meissner L, Dunaif A. The Galletly C, et al. Ghrelin and measures of satiety are altered in
independent effects of hyperandrogenaemia, polycystic ovary syndrome but not differentially affected by
hyperinsulinaemia, and obesity on lipid and lipoprotein diet composition. J Clin Endocrinol Metab 2004;89:3337–44.
profiles in women. Clin Endocrinol (Oxf) 1990;33:119–31. [54] Goverde AJ, van Koert AJ, Eijkemans MJ, Knauff EA,
[36] Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, Westerveld HE, Fauser BC, et al. Indicators for metabolic
et al, INTERHEART Study Investigators. Effect of potentially disturbances in anovulatory women with polycystic ovary
modifiable risk factors associated with myocardial infarc- syndrome diagnosed according to the Rotterdam consensus
tion in 52 countries (the INTERHEART study): case-control criteria. Hum Reprod 2009;24:710–7.
study. Lancet 2004;364:937–52. [55] Zhang HY, Zhu FF, Xiong J, Shi XB, Fu SX. Characteristics of
[37] Savva SC, Lamnisos D, Kafatos AG. Predicting cardiometa- different phenotypes of polycystic ovary syndrome based
bolic risk: waist-to-height ratio or BMI. A meta-analysis. on the Rotterdam criteria in a large-scale Chinese popula-
Diabetes Metab Syndr Obes 2013;6:403–19. tion. BJOG 2009;116:1633–9.
[38] Sam S. Obesity and polycystic ovary syndrome. Obes Manag [56] Lumeng CN, Saltiel AR. Inflammatory links between obesity
2007;3:69–73. and metabolic disease. J Clin Invest 2011;121:2111–7.
[39] Lim SS, Davies MJ, Norman RJ, Moran LJ. Overweight, obesity [57] Gonzalez F, Thusu K, Abdel-Rahman E, Prabhala A, Tomani
and central obesity in women with polycystic ovary M, Dandona P. Elevated serum levels of tumor necrosis
syndrome: a systematic review and meta-analysis. Hum factor in normal-weight women with polycystic ovary
Reprod Update 2012;18:618–37. syndrome. Metabolism 1999;48:437–41.
[40] Horejsi R, Möller R, Rackl S, Giuliani A, Freytag U, Crailsheim [58] Bañuls C, Rovira-Llopis S, Martinez de Marañon A, Veses S,
K, et al. Android subcutaneous adipose tissue topography in Jover A, Gomez M, et al. Metabolic syndrome enhances
lean and obese women suffering from PCOS: comparison endoplasmic reticulum, oxidative stress and leukocyte-
with type 2 diabetic women. Am J Phys Anthropol 2004;124: endothelium interactions in PCOS. Metabolism 2017;71:
275–81. 153–62.
42 M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 86 ( 20 1 8 ) 3 3– 43

[59] Escobar-Morreale HF, Botella-Carretero JI, Villuendas G, associated with insulin resistance and lipid accumulation
Sancho J, San Millán JL. Serum interleukin-18 concentra- product in women with polycystic ovary syndrome. Hum
tions are increased in the polycystic ovary syndrome: Reprod 2016;31:1347–53.
relationship to insulin resistance and to obesity. J Clin [77] Rocha ALL, Faria LC, Guimarães TCM, Moreira GV, Cândido
Endocrinol Metab 2004;89:806–11. AL, Couto CA, et al. Non-alcoholic fatty liver disease in
[60] Shroff R, Kerchner A, Maifeld M, Van Beek EJ, Jagasia D, women with polycystic ovary syndrome: systematic review
Dokras A. Young obese women with polycystic ovary and meta-analysis. J Endocrinol Invest 2017. https://doi.org/
syndrome have evidence of early coronary atherosclerosis. J 10.1007/s40618-017-0708-9 [Epub ahead of print].
Clin Endocrinol Metab 2007;92:4609–14. [78] Jones H, Sprung VS, Pugh CJ, Daousi C, Irwin A, Aziz N, et al.
[61] Eser B, Islimye Taskin M, Hismiogullari AA, Aksit H, Bodur Polycystic ovary syndrome with hyperandrogenism is
AS. The effects of IL-1A and IL-6 genes polymorphisms on characterized by an increased risk of hepatic steatosis
gene expressions, hormonal and biochemical parameters in compared to nonhyperandrogenic PCOS phenotypes and
polycystic ovary syndrome. J Obstet Gynaecol 2017;37: healthy controls, independent of obesity and insulin resis-
358–62. tance. J Clin Endocrinol Metab 2012;97:3709–16.
[62] Boulman N, Levy Y, Leiba R, Shachar S, Linn R, Zinder O. [79] Hossain N, Stepanova M, Afendy A, Nader F, Younossi Y,
Blumenfeld Z' Increased C-reactive protein levels in the Rafiq N, et al. Non-alcoholic steatohepatitis (NASH) in
polycystic ovary syndrome: a marker of cardiovascular patients with polycystic ovarian syndrome (PCOS). Scand J
disease. J Clin Endocrinol Metab 2004;89:2160–5. Gastroenterol 2011;46:479–84.
[63] Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive [80] Shi Y, Cui Sun X, Ma G, Ma Z, Gao Q, et al. Hypertension in
protein and other markers of inflammation in the predic- women with polycystic ovary syndrome: prevalence and
tion of cardiovascular disease in women. N Engl J Med 2000; associated cardiovascular risk factors. Eur J Obstet Gynecol
342:836–43. Reprod Biol 2014;173:66–70.
[64] Spritzer PM, Lecke SB, Satler F, Morsch DM. Adipose tissue [81] Elting MW, Korsen TJ, Bezemer PD, Schoemaker J. Preva-
dysfunction, adipokines, and low-grade chronic inflamma- lence of diabetes mellitus, hypertension and cardiac com-
tion in polycystic ovary syndrome. Reproduction 2015;149: plaints in a follow-up study of a Dutch PCOS population.
R219–27. Hum Reprod 2001;16:556–60.
[65] Athyros VG, Tziomalos K, Karagiannis A, Anagnostis P, [82] Orbetzova MM, Shigarminova RG, Genchev GG, Milcheva BA,
Mikhailidis DP. Should adipokines be considered in the Lozanov LB, Genov NS, et al. Role of 24-hour monitoring in
choice of the treatment of obesity-related health problems? assessing blood pressure changes in polycystic ovary
Curr Drug Targets 2010;11:122–35. syndrome. Folia Med (Plovdiv) 2003;45:21–5.
[66] Toulis KA, Goulis DG, Farmakiotis D, Georgopoulos NA, [83] Holte J, Gennarelli G, Berne C, Bergh T, Lithell H. Elevated
Katsikis I, Tarlatzis BC, et al. Adiponectin levels in women ambulatory day-time blood pressure in women with poly-
with polycystic ovary syndrome: a systematic review and a cystic ovary syndrome: a sign of a pre-hypertensive state?
meta-analysis. Hum Reprod Update 2009;15:297–307. Hum Reprod 1996;11:23–8.
[67] Gao L, Zhang Y, Cui Y, Jiang Y, Wang X, Liu J. Association of [84] Zimmermann S, Phillips RA, Dunaif A, Finegood DT,
the T45G and G276T polymorphisms in the adiponectin Wilkenfeld C, Ardeljan M, et al. Polycystic ovary syndrome:
gene with PCOS: a meta-analysis. Gynecol Endocrinol 2012; lack of hypertension despite profound insulin resistance. J
28:106–10. Clin Endocrinol Metab 1992;75:508–13.
[68] Zheng SH, Du DF, Li XL. Leptin levels in women with [85] Schmidt J, Landin-Wilhelmsen K, Brännström M, Dahlgren
polycystic ovary syndrome: a systematic review and a meta- E. Cardiovascular disease and risk factors in PCOS women of
analysis. Reprod Sci 2017;24:656–70. postmenopausal age: a 21-year controlled follow-up study. J
[69] Veldhuis JD, Pincus SM, Garcia-Rudaz MC, Ropelato MG, Clin Endocrinol Metab 2011;96:3794–803.
Escobar ME, Barontini M. Disruption of the synchronous [86] Cascella T, Palomba S, Tauchmanovà L, Manguso F, Di Biase
secretion of leptin, LH, and ovarian androgens in nonobese S, Labella D, et al. Serum aldosterone concentration and
adolescents with the polycystic ovarian syndrome. J Clin cardiovascular risk in women with polycystic ovarian
Endocrinol Metab 2001;86:3772–8. syndrome. J Clin Endocrinol Metab 2006;91:4395–400.
[70] Dambala K, Vavilis D, Bili E, Goulis DG, Tarlatzis BC. Serum [87] Pinola P, Puukka K, Piltonen TT, Puurunen J, Vanky E,
visfatin, vascular endothelial growth factor and matrix Sundström-Poromaa I, et al. Normo- and hyperandrogenic
metalloproteinase-9 in women with polycystic ovary syn- women with polycystic ovary syndrome exhibit an adverse
drome. Gynecol Endocrinol 2017:1–5. https://doi.org/10. metabolic profile through life. Fertil Steril 2017;107:788–95.
1080/09513590.2017.1296425 [Epub ahead of print]. [88] Mannerås-Holm L, Baghaei F, Holm G, Janson PO, Ohlsson C,
[71] Sun Y, Wu Z, Wei L, Liu C, Zhu S, Tang S. High-visfatin levels Lönn M, et al. Coagulation and fibrinolytic disturbances in
in women with polycystic ovary syndrome: evidence from a women with polycystic ovary syndrome. J Clin Endocrinol
meta-analysis. Gynecol Endocrinol 2015;31:808–14. Metab 2011;96:1068–76.
[72] Tang YL, Yu J, Zeng ZG, Liu Y, Liu JY, Xu JX. Circulating [89] Yildiz BO, Haznedaroğlu IC, Kirazli S, Bayraktar M. Global
omentin-1 levels in women with polycystic ovary syn- fibrinolytic capacity is decreased in polycystic ovary syn-
drome: a meta-analysis. Gynecol Endocrinol 2017;33:244–9. drome, suggesting a prothrombotic state. J Clin Endocrinol
[73] Polyzos SA, Mantzoros CS. Nonalcoholic fatty future dis- Metab 2002;87:3871–5.
ease. Metabolism 2016;65:1007–16. [90] Atiomo WU, Fox R, Condon JE, Shaw S, Friend J, Prentice AG,
[74] Gutierrez-Grobe Y, Ponciano-Rodríguez G, Ramos MH, Uribe et al. Raised plasminogen activator inhibitor-1 (PAI-1) is not
M, Méndez-Sánchez N. Prevalence of non-alcoholic fatty an independent risk factor in the polycystic ovary syndrome
liver disease in premenopausal, postmenopausal and poly- (PCOS). Clin Endocrinol (Oxf) 2000;52:487–92.
cystic ovary syndrome women. The role of estrogens. Ann [91] Atiomo WU, Condon J, Adekanmi O, Friend J, Wilkin TJ,
Hepatol 2010;9:402–9. Prentice AG. Are women with polycystic ovary syndrome
[75] Brzozowska MM, Ostapowicz G, Weltman MD. An associa- resistant to activated protein C? Fertil Steril 2000;74:
tion between non-alcoholic fatty liver disease and polycys- 1229–32.
tic ovarian syndrome. J Gastroenterol Hepatol 2009;24:243–7. [92] Tsanadis G, Vartholomatos G, Korkontzelos I, et al. Poly-
[76] Macut D, Tziomalos K, Božić-Antić I, Bjekić-Macut J, Katsikis cystic ovarian syndrome and thrombophilia. Hum Reprod
I, Papadakis E, et al. Non-alcoholic fatty liver disease is 2002;17:314–9.
M ET ABOL I SM CL IN I CA L A N D E XP E RI ME N TAL 8 6 ( 20 1 8 ) 3 3– 4 3 43

[93] Schachter M, Raziel A, Friedler S, Strassburger D, Bern O, [105] Meyer ML, Malek AM, Wild RA, Korytkowski MT, Talbott EO.
Ron-El R. Insulin resistance in patients with polycystic Carotid artery intima-media thickness in polycystic ovary
ovary syndrome is associated with elevated plasma homo- syndrome: a systematic review and meta-analysis. Hum
cysteine. Hum Reprod 2003;18:721–7. Reprod Update 2012;18:112–26.
[94] Domecq JP, Prutsky G, Mullan RJ, Sundaresh V, Wang AT, [106] Sprung VS, Atkinson G, Cuthbertson DJ, Pugh CJ, Aziz N,
Erwin PJ, et al. Adverse effects of the common treatments Green DJ, et al. Endothelial function measured using flow-
for polycystic ovary syndrome: a systematic review and mediated dilation in polycystic ovary syndrome: a meta-
meta-analysis. J Clin Endocrinol Metab 2013;98:4646–54. analysis of the observational studies. Clin Endocrinol (Oxf)
[95] Yildiz BO. Approach to the patient: contraception in women 2013;78:438–46.
with polycystic ovary syndrome. J Clin Endocrinol Metab [107] Okoroh EM, Boulet SL, George MG, Craig Hooper W.
2015;100:794–802. Assessing the intersection of cardiovascular disease, ve-
[96] Yousuf SD, Rashid F, Mattoo T, Shekhar C, Mudassar S, nous thromboembolism, and polycystic ovary syndrome.
Zargar MA, et al. Does the oral contraceptive pill increase Thromb Res 2015;136:1165–8.
plasma intercellular adhesion molecule-1, monocyte [108] Hart R, Doherty DA. The potential implications of a PCOS
chemoattractant protein-1, and tumor necrosis factor-α diagnosis on a woman's long-term health using data
levels in women with polycystic ovary syndrome: a pilot linkage. J Clin Endocrinol Metab 2015;100:911–9.
study. J Pediatr Adolesc Gynecol 2017;30:58–62. [109] Pierpoint T, McKeigue PM, Isaacs AJ, Wild SH, Jacobs HS.
[97] Lidegaard Ø, Nielsen LH, Skovlund CW, Skjeldestad FE, Mortality of women with polycystic ovary syndrome at
Løkkegaard E. Risk of venous thromboembolism from use of oral long-term follow-up. J Clin Epidemiol 1998;51:581–6.
contraceptives containing different progestogens and oestrogen [110] Wild S, Pierpoint T, McKeigue P, Jacobs H. Cardiovascular
doses: Danish cohort study, 2001–9. BMJ 2011;343:d6423. disease in women with polycystic ovary syndrome at long-
[98] McFarlane SI, Banerji M, Sowers JR. Insulin resistance and term follow-up: a retrospective cohort study. Clin
cardiovascular disease. J Clin Endocrinol Metab 2001;86:713–8. Endocrinol (Oxf) 2000;52:595–600.
[99] Emerging Risk Factors CollaborationWormser D, Kaptoge S, [111] Iftikhar S, Collazo-Clavell ML, Roger VL, St Sauver J, Brown Jr
Di Angelantonio E, Wood AM, Pennells L, et al. Separate and RD, Cha S, et al. Risk of cardiovascular events in patients
combined associations of body-mass index and abdominal with polycystic ovary syndrome. Neth J Med 2012;70:74–80.
adiposity with cardiovascular disease: collaborative analy- [112] Mani H, Levy MJ, Davies MJ, Morris DH, Gray LJ, Bankart J,
sis of 58 prospective studies. Lancet 2011;377:1085–95. et al. Diabetes and cardiovascular events in women with
[100] Talbott EO, Zborowski JV, Rager JR, Boudreaux MY, polycystic ovary syndrome: a 20-year retrospective cohort
Edmundowicz DA, Guzick DS. Evidence for an association study. Clin Endocrinol (Oxf) 2013;78:926–34.
between metabolic cardiovascular syndrome and coronary [113] Merz CN, Shaw LJ, Azziz R, Stanczyk FZ, Sopko G, Braunstein
and aortic calcification among women with polycystic ovary GD, et al. Cardiovascular disease and 10-year mortality in
syndrome. J Clin Endocrinol Metab 2004;89:5454–61. postmenopausal women with clinical features of polycystic
[101] Carmina E, Orio F, Palomba S, Longo RA, Cascella T, Colao A, ovary syndrome. J Womens Health (Larchmt) 2016;25:
et al. Endothelial dysfunction in PCOS: role of obesity and 875–81.
adipose hormones. Am J Med 2006;119:356.e1–6. [114] Wang ET, Cirillo PM, Vittinghoff E, Bibbins-Domingo K, Cohn
[102] Vryonidou A, Papatheodorou A, Tavridou A, Terzi T, Loi V, BA, Cedars MI. Menstrual irregularity and cardiovascular
Vatalas IA, et al. Association of hyperandrogenemic and mortality. J Clin Endocrinol Metab 2011;96:E114–8.
metabolic phenotype with carotid intima-media thickness [115] Solomon CG, Hu FB, Dunaif A, Rich-Edwards JE, Stampfer MJ,
in young women with polycystic ovary syndrome. J Clin Willett WC, et al. Menstrual cycle irregularity and risk for
Endocrinol Metab 2005;90:2740–6. future cardiovascular disease. J Clin Endocrinol Metab 2002;
[103] Kravariti M, Naka KK, Kalantaridou SN, Kazakos N, 87:2013–7.
Katsouras CS, Makrigiannakis A, et al. Predictors of endo- [116] de Groot PC, Dekkers OM, Romijn JA, Dieben SW,
thelial dysfunction in young women with polycystic ovary Helmerhorst FM. PCOS, coronary heart disease, stroke and
syndrome. J Clin Endocrinol Metab 2005;90:5088–95. the influence of obesity: a systematic review and meta-
[104] Armeni E, Stamatelopoulos K, Rizos D, Georgiopoulos G, analysis. Hum Reprod Update 2011;17:495–500.
Kazani M, Kazani A, et al. Arterial stiffness is increased in [117] Zhao L, Zhu Z, Lou H, Zhu G, Huang W, Zhang S, et al.
asymptomatic nondiabetic postmenopausal women with a Polycystic ovary syndrome (PCOS) and the risk of coronary
polycystic ovary syndrome phenotype. J Hypertens 2013;31: heart disease (CHD): a meta-analysis. Oncotarget 2016;7:
1998–2004. 33715–21.

You might also like