Introduction

Polycystic ovary syndrome (PCOS) is a very common and rather complex endocrine disorder with significant reproductive and metabolic impact . The overproduction of androgens seems to have a role in its development, hence not just being related to the well-known clinical manifestations of the syndrome and its long-term consequences.

There have been many attempts to explain the development of PCOS. In its origin, a suspected genetic basis has been proposed. The more recent evidence indicates that PCOS is probably a complex polygenic disorder with strong environment influences, notably those associated with obesity.

Genomewide association studies (GWAS) have implicated a significant number of genes in the development of PCOS. These include genes for the gonadotropin receptors LHCGR and FSHR, the beta subunit of follicle-stimulating hormone (FSH), insulin receptor (INSR), YAP1, DENND1A , and THADA . After GWAS, the implicated loci and genetic variants need to have their biologic effect and etiologic significance ascertained by functional studies and this process is still going on without illuminating results. Analyses of rarer genetic variants have produced results that on the whole are even more influential than those with higher allelic frequencies identified in GWAS.

Many genes have been suspected (and confirmed) and are referred to as the most plausible candidate genes. These include genes related to gonadotropin secretion, or action, ovarian folliculogenesis, steroidogenesis, insulin secretion or action, and adipose tissue function, among others . These candidate genes will still have to be fully sequenced in appropriately phenotyped PCOS cases and in appropriately dimensioned patient series in order to identify all coding variants that may be contributing to PCOS. Nevertheless, it is already evident that the androgen production by the ovaries and its action through the androgen receptor (AR), together with all of the modulators contributing to the generation of a hyperandrogenic milieu, are at the center of the genetic background that leads to PCOS development.

Under the concept of polycystic ovary syndrome (PCOS) as a complex genetic trait, adrenal androgen (AA) production and excess is also to be considered a relevant factor for the development of the disorder. Data from girls with premature adrenarche and patients with 21-OH deficient CAH support the role of AA excess in the development of the PCOS phenotype in many patients.

In fact, in many PCOS women, a generalized adrenocortical hyperresponsivity is present, particularly in those with overt AA excess, similar to the ovarian hyperfunction seen in the disorder .

Whatever be the origin of androgens, early androgen excess, in particular, prenatal androgen exposure, has been emerging as a possible trigger for PCOS occurrence later in life, being an example of “Developmental Programming” through which hormones acting during the embryonic period contribute to the development of diseases, and in this case PCOS. The evidence, despite still not being absolute, is compelling and seems to guarantee that this is one of the most important initiating factors for PCOS development.

PCOS clinical aspects

PCOS clinical aspects involve many different symptoms and signs going from hirsutism to menstrual irregularities, from insulin resistance to obesity and finally also infertility. One should also not forget the reduced self-esteem, anxiety, and depression that so frequently affect these women. PCOS is associated with increased metabolic morbidity, including increased adiposity, glucose intolerance, and type 2 diabetes mellitus (T2DM). It has also been suspected to have significant cardiovascular consequences.

Since the adoption of the Rotterdam criteria for the definition of PCOS , an added difficulty in identifying its etiology has resulted from the understanding that more than one phenotype exist, taking into consideration the criteria of the new classification.

PCOS is identified by the presence of 2 of the following 3 possible criteria :

• •
  

  polycystic ovaries confirmed by ultrasonography,

  
  
• •
  

  hyperandrogenism (either clinical or biochemical), and

  
  
• •
  

  ovulatory dysfunction (defined clinically by significant menstrual irregularities and confirmed by biochemical determination of progesterone levels in the supposed luteal phase in women that maintain menstruation cycles).

Consequently, there are currently four recognized phenotypes of PCOS,

• (1)
  

  hyperandrogenism + oligoanovulation + polycystic ovarian morphology;

  
  
• (2)
  

  hyperandrogenism + oligoanovulation;

  
  
• (3)
  

  hyperandrogenism + polycystic ovarian morphology; and

  
  
• (4)
  

  oligoanovulation + polycystic ovarian morphology,

each with different long-term health and metabolic implications.

Investigating those different phenotypes has made clear to the scientific community that PCOS is not one single disorder but a group of disorders probably with different etiologies or different weights of the diverse etiologic factors.

Besides, definitions have evolved to encompass even cases without hyperandrogenism and at the same time attempted to expand the possibility of identification of the syndrome, to ages that are not restricted to the span between puberty and menopause, hence posing new challenges in the etiologic understanding of PCOS when ovaries are nonfunctioning and sex steroids, including androgens, are at residual levels.

Polycystic ovary syndrome (PCOS) is one of the most common endocrine disorders, affecting approximately 5%–10% of women of reproductive age. If one considers all the phenotypes resulting from Rotterdam, this prevalence increases to 20% of women in that age span . So, PCOS is considered the most important cause of androgen excess and anovulation in women between 18 and 40 years of age.

Hyperandrogenism is considered to be one of the most important symptoms of PCOS because elevated concentrations of androgens, including testosterone, androstenedione, and dehydroepiandrosterone sulphate, are observed in the majority of patients .

Hyperandrogenism is mainly originated in the ovary, with a possibly significant contribution from the adrenal gland and also the adipose tissue .

It has also been shown that abnormally low concentrations of serum SHBG are observed in women with PCOS and that this contributes to the hyperandrogenic symptoms such as hirsutism and acne . A recent meta-analysis involving 62 articles analyzing the potential relationship between serum SHBG concentrations and the occurrence of PCOS demonstrated that lower serum SHBG concentrations were associated with higher risk of PCOS . These results clearly indicate that a down-regulation of SHBG expression plays an important role in the pathogenesis of PCOS and suggests that serum SHBG concentrations could be used as a biomarker or therapeutic target for the diagnosis and treatment of PCOS.

Some experts have argued that hyperandrogenism should be required for a diagnosis of the polycystic ovary syndrome because it best identifies women at risk for coexisting metabolic conditions. As a group, women with only ovulatory dysfunction and polycystic ovaries have lower cardiometabolic risk than women with classic polycystic ovary syndrome .

The next most important sign in PCOS, ovulatory dysfunction, is typically indicated by unpredictable menses that occur at less than 21-day or greater than 35-day intervals. However, having regular menses that occur every 21 to 35 days do not confirm normal ovulatory function in women with hyperandrogenism since 15 to 40% of women with hyperandrogenism and regular menses have anovulatory dysfunction .

Androgens in PCOS

Androgen excess has been identified in most cases of PCOS, a disorder that was in the past considered an Androgen Excess disorder until the consideration of at least one phenotype without elevated androgens and without hirsutism, by the Rotterdam criteria.

Androgen excess in women with PCOS may be primarily ovarian and/or adrenal in origin. Androgen excess has been studied both clinically by analyzing the levels of different androgens in individual patients and in populations and also by dissecting the enzymatic pathways that leads to their production and their effects in different tissues. Various intracellular pathways, mostly implicated in steroidogenesis and various receptors and postreceptor pathways that are related to androgenic action, have been studied. We now know a lot more than we used to, and, in fact, in recent years even new androgens have been identified and their importance in PCOS is well established. Still, we do not know everything.

Clinically, serum total testosterone and free testosterone, or the free androgen index (FAI-calculated as Total Testosterone/SHBG x100), in conjunction with SHBG, are usually evaluated.

Additionally, androstenedione, DHEA and DHEA-s, 17OH progesterone, and dehydrotestosterone (DHT), as well as the newer androgens 17 keto-testosterone and 17 keto-DHT , are also important for the interpretation of androgen excess conditions and should be included in an evaluation of these patients, with research purposes. All of these are, in fact, needed to the best identification of hyperandrogenism in PCOS and understanding of its physiological role, including the etiopathogenetic influence. Not performing a complete androgen assessment only contributes to not being able to understand the correlation between androgen levels and the clinical symptoms.

Androgens, in women, are produced mainly in the ovaries, but the adrenal and the adipose tissue contributions to the complete picture of circulating androgens should not be forgotten. Androgen synthesis involves several enzymes like the microsomal P450c17 with both 17α-hydroxylase and 17,20-lyase activities. A dysfunction in steroidogenesis, mainly in the androgenic pathway, has been pointed out as having a role in PCOS development .

The ovaries, on the other hand, also seem to be hyperresponsive to luteinizing hormone (LH), and since LH is increased in these patients, its effect is even higher . Increased LH secretion seems to result from a hypothalamic defective response to the feedback inhibition of gonadotropin-releasing hormone (GnRH) secretion. This feedback inhibition normally results from the estradiol and progesterone circulating levels, and in PCOS, this process is impaired as a result of the excessive androgen concentrations .

Hyperinsulinemia, a consequence of the insulin resistance that frequently is observed in PCOS, also contributes to the androgen excess either directly, by stimulation of the theca cells, or indirectly, by inhibiting hepatic synthesis of SHBG, consequently increasing the free (active) fraction of circulating androgens.

Insulin also acts at the adrenal level in synergy with ACTH, thus also increasing the secretion of adrenal androgens . It has been demonstrated that adrenal androgens are increased in 20%–36% of PCOS cases. The importance of this adrenal androgen’s excess in PCOS has not been completely elucidated. It has been postulated that it may be important by one of two possibilities: Either the production of adrenal androgens starts precociously, a phenomenon known as early adrenarche, and this precocious hyperandrogenic milieu leads to PCOS; or adrenal androgens overproduction is part of a generalized androgen synthesis occurring in all steroidogenic tissues, in PCOS.

Adrenal androgens can, in fact, be secreted prior to puberty as demonstrated in cases of CAH and NCAH but also in the cases of premature adrenarche in which it was demonstrated that 50% come to develop PCOS .

It has also been reported that P450c17 activity is augmented in the adrenals of PCOS patients who have elevated adrenal androgens. It was suggested that an augmented cortisol metabolization would cause reduced cortisol levels and reduce feedback on the hypothalamus-pituitary axis, leading to an increase in ACTH and consequently to hyperproduction of adrenal androgens . However, ACTH is normally not increased in PCOS women, and so, it is more plausible that there is an overresponsiveness of the zona reticularis of the adrenal cortex to normal levels of ACTH in women with PCOS having increased adrenal androgens .

We may conclude that the evidence of a possible participation of adrenal androgens in the development of PCOS was demonstrated in cases of inherited diseases running with adrenal originated hyperandrogenism (CAH and NCAH) as well as in cases in whom an exaggerated adrenarche with elevated adrenal androgens precedes the development of PCOS, thus allowing us to suspect that one situation is causing the other.

Finally, the influence of excessive adipose tissue on androgen levels has also been recognized, and, in fact, obesity is associated with more severe phenotypes while body weight reduction, at the level of 5%, is associated with androgen levels reduction and an overall improvement of their clinical manifestations .

Pathophysiology of PCOS

First, we need to take into consideration that there are intrinsic abnormalities of ovarian steroidogenesis, namely, resulting from the impairment of follicular development with a consequent hypertrophy of the theca layer and underdevelopment of the granulosa. The increased activity of P450c17 in the theca and the reduced P450c19 (aromatase) activity in the granulosa layer of the ovarian follicle are responsible for an excessive androgen production as well as for ovulatory dysfunction. Intrinsic abnormalities of ovarian steroidogenesis have been clearly demonstrated: cultured ovarian theca cells from women with the polycystic ovary syndrome secrete excess androgens and precursors , and women with the polycystic ovary syndrome have exaggerated ovarian steroidogenic responses to gonadotropin stimulation . A recent study suggested that increased expression of a DENND1A splice variant drives a polycystic ovary syndrome-like steroidogenic phenotype in theca cells .

The majority of women with PCOS often have insulin resistance and consequently hyperinsulinemia. Hyperinsulinemia contributes to hyperandrogenemia in several ways: it augments LH-stimulated androgen production by ovarian theca cells, it potentiates corticotropin-mediated adrenal androgen production, and in turn inhibits hepatic synthesis of sex hormone-binding globulin (SHBG), which increases free testosterone levels. Hyperinsulinemia probably also affects gonadotropin secretion. Consequently, hyperandrogenism is aggravated by the hyperinsulinism resulting from obesity and from other causes of insulin resistance, which, beyond many other effects, induce androgen production at the ovaries and possibly at the adrenal level as well.

Beyond its participation in insulin resistance the adipose tissue also contributes importantly to the etiopathogenesis of PCOS, namely, by the secretion of some adipokines, like leptin, or by the reduced secretion of others, like adiponectin. Besides, the adipose tissue is also one of the most important organs in converting androgenic precursors like DHEA or androstenedione to stronger androgens like testosterone and also by converting androgens to estrogens, leading to a noncycling estrogenic milieu that disrupts the normal functioning of the hypothalamic-pituitary-ovary axis.

Finally, the hypothalamic pituitary axis is also implicated in the pathogenesis of PCOS being associated with an increased androgen production that, besides the above described mechanisms, also results from the increased LH production in conjunction with a relative deficiency in FSH production. In PCOS, a persistently rapid GnRH pulsatility is observed and is responsible for the increased LH and the imbalance between LH and FSH secretion having the consequence of an excessive production of androgens.

Notably, androgens seem to play a role in this alteration of GnRH pulsatility and the consequent increase of LH secretion. High levels of androgens interfere with the hypothalamic sensitivity to the normal negative feedback from estrogens and progesterone, hence being major contributors to the increase in GnRH pulse frequency . Reciprocally, GnRH secretion with rapid pulse frequency favors increased LH secretion, which acts mostly on the theca cells stimulating the production of androgens . Simultaneously, the rapid pulse frequency of GnRH results in a relative decrease of FSH, which leads to poor development of the granulosa layer of the ovarian follicles with less aromatase activity and less conversion of androgens to estradiol.

In summary, since ovarian steroidogenesis requires gonadotropin stimulation, LH is a key factor in the hyperandrogenemia of the polycystic ovary syndrome . Progesterone is the primary regulator of GnRH pulse frequency; however, in polycystic ovary syndrome, the GnRH pulse generator is relatively resistant to the negative feedback effects of progesterone . This resistance to progesterone negative feedback appears to be mediated by androgen excess since it is reversed by the androgen-receptor blocker flutamide . The inhibition of progesterone feedback is linked with high GnRH pulse frequencies, which favors the production of LH and limits the production of FSH, which promotes androgen secretion and interferes with normal follicular development.

The androgenic hypothesis in the pathogenesis of PCOS

Many studies on animal models of PCOS as well as congenital or early infancy situations of androgen excess in women raised the hypothesis that early exposure to androgens may be the most important cause leading to the occurrence of PCOS.

An excessive synthesis and secretion of androgens, acting on AR, may have an important role in the genesis of polycystic ovary syndrome (PCOS). To analyze this hypothesis, it is very important to know the precise locations of AR as well as the consequences of their activation in those different locations. The interest in the role of androgens on the PCOS development goes far beyond a mere scientific knowledge. It may prove important in the delineation of targeted treatments of the syndrome.

Data from girls with premature adrenarche and patients with 21-OH deficient CAH support the role of AA excess in the development of the PCOS phenotype: adrenal diseases with elevated production of adrenal androgens have been shown to produce an ovarian phenotype that is similar to PCO (polycystic ovaries with increased volumes and thecal/interstitial hyperplasia) and clinical manifestations that are similar to PCOS .

The same was observed in women who are subjected to high levels of androgens as a result of exogenous testosterone treatment in female-to-male transsexuals . These observations are all in favor of a role for androgens in the development of PCOS.

An increased activity of the AR also seems to contribute to the development of PCOS. The AR is encoded by a gene located at Xq11-q12 . Exon 1 of the AR gene contains a variable length CAG repeat polymorphism . A smaller number of CAG repeats has been associated with hirsutism , as well as ovarian hyperandrogenism in women .

Association of shorter CAG repeats with PCOS suggests that inherited alteration in androgen sensitivity may contribute to PCOS . In addition, a decreased number of AR gene CAG repeats (which is associated with increased androgen sensitivity) could explain the normal serum androgen levels in women with PCOS that seem to have only clinical hyperandrogenism (hirsutism) .

The effect of antiandrogens

Flutamide, an antagonist of the AR, has been many times reported to have a good therapeutic effect in women with PCOS, notably by decreasing hirsutism . However, it was also observed that PCOS patients when subjected to treatment with flutamide ameliorated their menstrual cycle regularity as well as their ovulatory function , thus confirming that lowering androgen levels not only improves what is believed to be the direct androgenic impact on the clinical manifestations of hyperandrogenism but also influences all of the components of the syndrome, including the ovulatory dysfunction and infertility, thus pointing out that androgens have an etiologic role in the development of the entire clinical syndrome.

Studies on animal models

Even researchers not very much convinced by the so-called animal models of PCOS have to acknowledge that they constitute an unique opportunity to study aspects of the pathogenesis of PCOS in ways that are not possible to do in humans . Many animal models of PCOS have been created by the administration of androgens inducing a situation of hyperandrogenism. It became possible to analyze if androgens were only responsible for the hyperandrogenic symptoms or also affected other aspects of PCOS, in particular, the ovarian reproductive function, the secretion of estrogens and progesterone, and finally if they contributed to metabolic alterations similar to those observed in humans .

In particular, the administration of androgens during pregnancy seems to be able to produce a syndrome with significant similarities to human PCOS. Examples of androgens that have been used in this context are testosterone and DHT. The obtained effects vary, not only according to the used androgen but also the timing of their administration. When done in the most adequate dosage and timing, prenatal androgen administration produces hormonal, metabolic, and fertility alterations that resemble PCOS . Animals in these studies not only have increased androgen levels but also elevated LH and subfertility . Many authors described the occurrence of polycystic ovarian morphology with numerous arrested antral follicles. In other cases, alteration of ovarian cyclicity or reduced fertility was demonstrated. The typical metabolic consequences of PCOS were also observed in animal studies of PCOS models.

Like in humans, AR antagonists have been used in these animal models and seem to confirm the etiopathogenic role of androgens in the development of PCOS. Also in animals, antiandrogens were able to prevent or correct some of the signs of PCOS. In some models with anovulation, for instance, an improved LH secretion leading to ovulation was observed . In another example using a mouse model of PCOS, flutamide improved estrous cycling and reduced the number of cyst-like follicles in the ovaries .

All of these studies seem to corroborate that androgens play a significant role in the development of PCOS.

Another model of research used androgen receptor knockouts (ARKO) and demonstrated that without the AR, PCOS could not be induced in female rodents exposed to androgen administration .

In what concerns where do the androgens act to produce a PCOS-like phenotype, some elegant studies have been performed, knocking down the receptors in different locations. When the knocking down of the AR was only induced in the granulosa cells of the ovary, or only in the theca cells, the development of features of PCOS was still observable, seeming to prove that the ovary is not the major target of androgens as inducers of PCOS . Surprisingly, knocking out the AR in the brain prevented the development of most reproductive and metabolic traits that are typical of PCOS , thus pointing out the brain as a major location for androgens to induce the development of PCOS.

Similarly, to what is observed in women with PCOS, in the rodent’s PCOS models, there is frequently an increase in LH to FSH ratio and an increased LH pulse frequency .

However, GnRH neurons, which are known to be the major regulators of LH and FSH secretion, do not express AR . On the other hand, they are expressed in the kisspeptin-neurokinin B-Dynorphin “KNDy” system in the hypothalamus, which is involved in the hypothalamic control of GnRH secretion . It was observed that there are increased kisspeptin levels in at least some PCOS patients , while its expression and neuronal networks have been observed to be altered in hyperandrogenized animal PCOS models . Adding to this, a study treating PCOS patients with a neurokinin-3 (NK3) receptor antagonist reduced LH (and LH pulse frequency) as well as testosterone blood levels, thus confirming this pre-GnRH neuronal system as important in PCOS development as well as a treatment target .

Another important aspect concerning the hypothalamic importance in the development of PCOS, namely, of its metabolic alterations, comes from the observation that leptin, a secretory product of adipose cells, which is increased in PCOS, acts on the POMC and neuropeptideY/Agouti-related peptide (NPY/AgRP) neurons . In hyperandrogenic-induced PCOS models, POMC mRNA and fiber projections were reduced, while NPY/AgRP cell number and fiber projections were augmented, thus establishing a link between hyperandrogenic-induced PCOS metabolic alterations, such as obesity and the hypothalamic dysfunction, which causes the alterations in gonadotropins production . Besides, in an extension of these studies, flutamide treatment was able to prevent at least the NPY/AgRP neurons’ changes .

All of these studies seem to prove that androgens, acting through centrally located ARs, play an important role in the development of the gonadotropin and metabolic alterations observed in PCOS.

Androgens’ action at the hypothalamus also affects the adipose tissue and participate in what can be defined as an androgen-brain-adipose tissue axis (ABAA), leading to the development of PCOS. Visceral adipose tissue, expressed as intra-abdominal fat mass, is increased in women with PCOS and positively correlated with circulating androgen levels . On the reverse, androgens in excess affect adipose tissue function in PCOS altering the normal production of adipokines. Recent studies revealed that increased leptin concentration frequently occurs in PCOS patients. This can be connected with the occurrence of impaired glucose tolerance and obesity. The fact that leptin is increased in PCOS and yet patients are frequently obese has been attributed to leptin resistance, similarly to what occurs with insulin resistance and hyperinsulinemia . Nevertheless, well-designed studies concerning the role of leptin in PCOS pathogenesis are still needed.

On the contrary, adiponectin has been demonstrated to be lower in women with PCOS , while its administration or its increment through the transplantation of brown adipose tissue in PCOS animal models may be able to reverse PCOS features . Increase in adiponectin prevented mice subjected to androgen administration to develop metabolic PCOS traits .

Prenatal and early postnatal androgen exposure

Prenatal androgen exposure in different animal models was able to recapitulate several clinical traits that are used to define PCOS in women. This has led to the hypothesis that prenatal androgen excess might contribute to the origin of PCOS development . The potential mechanism via which androgens would lead to PCOS development might imply the increment of GnRH and LH secretion, directly influencing steroidogenic enzymes toward the production of androgens or increasing insulin secretion while decreasing insulin responsiveness .

However, for the transposition of these results to women, it would be important to determine what could be the source of intrauterine elevation of androgens in the case of female human embryos. The androgenic milieu obtained in animal models by exogenous administration is virtually impossible to replicate in human pregnancies even when the mother has hyperandrogenemia as in the case of PCOS women. To complicate the parallelism with the animal models, the human placenta through its high levels of aromatase activity normally impedes the mother’s hyperandrogenism to be transposed to the fetus. Contrarily to this, one study reported that the placenta of PCOS women has subnormal aromatase activity , thus being more predisposed to allow the transfer of androgens from mother to daughter and, so, justify to some level the occurrence of hyperandrogenism in those pregnancies and the consequent “programming” to develop PCOS in adulthood . Placental insufficiency might be another possible justification allowing the occurrence of intrauterine hyperandrogenic environment in the daughters of PCOS mothers.

Besides to prenatal exposure to an hyperandrogenic environment, neonatal and prepubertal increments in androgen levels may also result in alterations that contribute to the development of PCOS, namely, insulin resistance and consequently hyperinsulinemia and obesity, which are well-known etiological factors or contributors to its severity .

Several molecular mechanisms explaining the relationship between hyperandrogenemia and insulin resistance have been proposed, but the fact is that a full conclusion has not been reached yet .

Androgens are postulated to be responsible for central obesity in hyperandrogenic syndromes by affecting the adipose cells, its hormonal and adipokine production, and also by acting synergically with insulin and glucocorticoids . Besides, in women, androgens contribute to obesity acting on specific hypothalamic nuclei and negatively on the brown adipose tissue, increasing appetite, food intake, and decreasing energy consumption .

Hyperandrogenemia is also one of the main contributors to anovulation in PCOS by their direct influence in the folliculogenesis. Androgen receptors are observable in the ovarian follicles before FSH receptors, and hyperandrogenemia has been implicated in precocious follicle growth and for exaggerated numbers of follicles initiating development at each cycle . In PCOS follicles, androgens act together with LH and insulin inhibiting granulosa development and aromatase activity, thus affecting estrogen production.

It seems undeniable that females that develop PCOS underwent through some period of previous androgen excess .

Metabolic effects of elevated androgens

The importance of PCOS is not limited to the reproductive consequences of the syndrome as metabolic alterations and their consequences are frequently observed. Obesity affects the majority of PCOS cases, and hyperinsulinemia, which is known to occur in both conditions, is clearly more significant when both are present simultaneously .

Insulin resistance is the hallmark of the metabolic alterations in PCOS. Women with PCOS often present with hyperinsulinemia, consequent to this insulin resistant state independently of being lean or obese, although obesity is certainly an important factor contributing to this etiopathogenic factor. Insulin resistance explains the risk of T2DM occurrence in PCOS women .

This is particularly relevant in PCOS cases with the classic phenotypes, who have metabolic consequences clearly more significant than PCOS cases without hyperandrogenism . PCOS cases without hyperandrogenism, in fact, have a much smaller risk of developing manifestations of the metabolic syndrome or type 2 diabetes .

PCOS also presents alterations of the blood lipids with increased triglycerides and decreased HDL cholesterol levels, an alteration that is more pronounced in cases with higher androgens .

Hyperandrogenism was also shown to correlate with other vascular risk factors such as the carotid intima-media thickness .

Altogether, most evidence seems to support the fact that elevated androgen levels contribute significantly to the metabolic alterations and expected cardiovascular risk associated with PCOS.

The influence of hyperandrogenism in the development of the metabolic alterations takes place affecting peripheral organs like the liver, the muscles, the pancreas, the adipose tissue, as well as some CNS centers related to appetite and energy expenditure.

At the hepatic level, androgens may be the most important contributor to insulin resistance, since the use of antiandrogens has been shown to reduce this hepatic insulin resistance in PCOS .

At the skeletal muscle, the influence of hyperandrogenism in women is related to the reduction of insulin sensitivity.

Androgen excess also affects β cell function in women with PCOS . Pancreatic β cells were demonstrated to express androgen receptors , suggesting that androgens might have a direct impact on pancreatic function.

It seems to be widely accepted that excess androgen in women leads to an increment and altered function of visceral adipose tissue. Androgens are important regulators of lipolysis as well as of lipogenesis .

Hyperandrogenic women exhibit a fat distribution that bears some similarities to the male distribution with important visceral fat deposition, namely, at the liver (NASH) and at peritoneal and mesenteric localizations. Simultaneously they reduce the subcutaneous fat accumulation, to some extent, which is typical of females . In female-to-male transsexuals who are treated with testosterone, this androgen induces the same metabolic alterations that are observed in PCOS women and so help to demonstrate the contribution of androgens to these alterations in adipose tissue deposition that are observed in hyperandrogenic phenotypes of the syndrome . Contrarily, it was observed that the treatment of PCOS cases with flutamide, an antiandrogenic drug, produced a reversal of this visceral predominant adipose accumulation . The importance of these alterations in adipose tissue localization results from the fact that visceral fat is more associated with the metabolic syndrome and increased cardiovascular risk .

The effect of androgens on adipocytes consists of an increment of their size , which, associated with macrophagic invasion, causes significant alterations in cytoquine production , and a reduction of adipocyte differentiation .

Androgens interfere significantly with the adipokine’s secretion by the adipose tissue. Adiponectin, which is an adipokine with insulin-sensitizing activity, is reduced by androgens in PCOS women , and probably, this reduction is a significant contributor to the occurrence of insulin resistance in PCOS.

Finally, androgen excess’ influences on metabolic alterations can also result from its actions on hypothalamic centers . The largest regulation of appetite takes place in the arcuate nucleus through the reduction of POMC or the activation of AgRP expressing neurons. Besides AgRP, AgRP neurons also express NPY, which is also orexigenic. AgRP/NPY and POMC neurons have receptors for both insulin and leptin , but also AR .

Exposure to androgens during intrauterine life increases AgRP/NPY neurons in adult females contributing to obesity through increased food ingestion . Simultaneously, there may be a reduction of POMC neurons, which will end up having the same effect . Besides, androgens reduce the sensitivity to leptin, and consequently, leptin loses its capacity to contribute to weight loss .

Interestingly, NPY neurons have another important role, namely, by constituting the interface between the regulation of body weight and fertility by indirectly signaling to the GnRH neurons, through kisspeptin secreting neurons, subsequently, affecting LH pulsatile secretion .