Introduction

Polycystic ovarian syndrome (PCOS) is one of the most common diseases of the endocrine system in women of reproductive age. The incidence ranges from 5 to 10% in different studies [1–3]. About 70% of women who have anovulation and normal serum gonadotropin levels suffer from the syndrome [2], while 50% of those with PCOS have increased body weight or are obese [4]. The clinical and laboratory picture of PCOS is characterized by a combination of symptoms such as oligo/amenorrhea, clinical or laboratory evidence of elevated androgens and ultrasound imaging of polycystic ovaries [3]. According to ESHRE/ASRM [3], women with PCOS were separated in the following groups: (a) women with hyperandrogenism, oligoanovulation, and PCO (full-blown syndrome); (b) women with hyperandrogenism and oligoanovulation, but without the ultrasound appearance of PCO; (c) women with hyperandrogenism and PCO (ovulatory PCOS); and (d) women with oligoanovulation and PCO (non-hyperandrogenic PCOS).

The metabolic findings of PCOS imply increased insulin resistance, hirsutism, increased androgens levels, dyslipidaemia, and hypertension, factors which increase the incidence of delayed cardiovascular problems [4]. Thus, about 40% of middle-aged women with PCOS present diabetes mellitus type II or glucose intolerance [4].

The PCOS is often accompanied by infertility requiring ovulation induction by using clomiphene citrate and/or gonadotropins [4]. Pregnancy is a predisposing factor of increased insulin resistance [4]. It is estimated that 25–70% of women with PCOS show an increase in insulin resistance and have an increased risk of developing complications during pregnancy, especially gestational diabetes [5]. Indeed, according to the Barker hypothesis, there is an in utero fetal programing of fetal nutrition and of the endocrine and metabolic environment in which the fetus develops, a fact that explains why PCOS complications might affect the fetus [6]. Many women with PCOS undergo assisted reproduction techniques in an effort to achieve pregnancy, fact which increases the incidence of multiple pregnancies with all the consequences caused by them compared to the singleton pregnancies [7]. Although, literature review about PCOS and pregnancy outcomes was previously conducted [8–17], it is still uncertain, however, to what extent PCOS itself affects pregnancy outcome in general and perinatal outcome.

This mini-review contains the relevant literature search for publications in PubMed and relevant references from the initially identified articles between the years 1960 and 2010. The used keywords included the terms “PCOS”, “complication”, “pregnancy”, “metformin”, “diabetes mellitus”, “spontaneous abortion”, “miscarriage”, “congenital anomalies”, “preeclampsia”, and hypertension”. The purpose of the review is to present the complications, the perinatal morbidity and mortality of women with PCOS, and the possible therapeutic role of metformin in those cases. All data that did not use the English language were excluded. From the relevant articles on the field, the potential role of PCOS in such conditions is presented.

Complications of pregnant women with PCOS

Complications of the first trimester of pregnancy which were studied in women with PCOS were the miscarriages and fetal congenital abnormalities. It has been suggested that PCOS is associated with an intrauterine inflammatory response characterized by elevated white blood count and CRP which impedes implantation [18–21]. The obesity that characterizes the syndrome also appears to play an important role in the pathophysiological pathway [5, 22, 23].

Some studies, and particularly that of Sir-Petermann et al. [24–27], have indicated no relationship of PCOS to the presence of congenital anomalies. However, a metaanalysis in 2006 revealed a 1.6–1.7 relative risk [26]. Sir-Petermann et al. [27] reported that the incidence of birth defects among pregnant women with PCOS and normal pregnancies was not found to be statistically different (OR 0.70; 95% CI 0.11–4.39).

No clear evidence is found whether there is an increased risk of spontaneous abortion rates in women with PCOS. Some authors have reported that women with PCOS have a 33% chance of spontaneous abortion [28, 29]. Both insulin resistance and increased body weight are associated with PCOS. In a recent study, an increased incidence of miscarriage rates is found in women with increased insulin resistance [30]. Furthermore, an increased incidence of miscarriages in pregnant women with increased body weight has been shown [31]. It has also been demonstrated that miscarriage rates are increased in women with elevated serum levels of LH. More specifically, high serum levels of LH, abnormal production of progesterone, high serum levels of androgens, hyperomokysteinaimia, and increased levels of glucodelins and IGFBP-2 appear to be involved in the pathogenesis of the syndrome [24, 28, 32–36]. Nevertheless, when gonadotropins are used in assisted reproduction techniques in women with the syndrome, no increased spontaneous abortions rates have been mentioned [32, 37]. On the contrary, other studies did not find increased rates of spontaneous abortions in women with PCOS. In those studies, however, women without PCOS were compared to women with either unique ultrasound findings of polycystic ovaries [29, 38, 39] or obese women with or without PCOS [40, 41] or women with high levels of luteal phase LH and testosterone [42].

Complications associated with PCOS in the second to third trimester of pregnancy are the impaired glucose tolerance, gestational diabetes, arterial hypertension, and preeclampsia [43–54].

Holte et al. [43] showed in 1998 that the prevalence of gestational diabetes was significantly higher among PCOS versus non-PCOS women. In the same metaanalysis, it has been shown that women with history of PCOS and no statistical significant difference in BMI and/or age had increased incidence of gestational diabetes [43]. Mikola et al. [44] also indicated an increased risk for gestational diabetes primarily because of obesity than the PCOS itself. Obesity is associated with elevated insulin levels and glucose intolerance. For this reason, in the same study one-third of the patients with PCOS were overweight. Alshammari et al. [13] in another study showed that women with both PCOS and gestational diabetes mellitus have an increased risk of obstetric complications such as pregnancy-induced hypertension/preeclampsia, preterm labor, and need for phototherapy treatment of newborns for hyperbilirubinemia. In a recent metaanalysis, women with PCOS had an increased risk of gestational diabetes [45]. However, the metaanalysis did not count confounding factors such as age, gravida, and multiple pregnancies, which are correlated with gestational diabetes.

Studies have reported increased risk of preeclampsia [46], hypertension [47, 48], and gestational diabetes [49–52] in women with anoovulating cycles. Regarding arterial hypertension during pregnancy in women with PCOS, it has been shown that there is a statistically significant difference in systolic, diastolic, and mean arterial pressure in these women compared to normal women. This finding has also been shown by Hu et al. [53] after measuring carotid stiffness in pregnant women with PCOS. All previous studies and a recent metaanalysis have shown the correlation of arterial hypertension during pregnancy in women with PCOS [47, 54]. In the same metaanalysis, a 3.5 times higher risk for preeclampsia and 1.75 times for preterm birth have been shown in pregnant women with PCOS [54]. It should be mentioned that the rates of iatrogenic preterm labor are increased in women with PCOS [44, 55]. However, Legro et al. [12] showed that birthweight in families with PCOS has no substantive association either with reproductive or with metabolic abnormalities in women with PCOS or their relatives.

Regarding the manner in handling childbirth, older studies showed high rates of cesarean sections [44, 55]. A metaanalysis by Boomsa et al. [54] in 2008 showed that women with PCOS had a 1.5 times higher risk of cesarean section. But when the control group was matched for age and BMI, no statistically significant difference was found regarding normal delivery, forceps use or cesarean section [54].

Newborns of women with PCOS had a 2.31 times higher risk of admission to intensive care unit for neonates and three times higher perinatal mortality than neonates of women without PCOS [54]. Perinatal morbidity of these neonates could be explained by prematurity and intrauterine growth retardation, and the IVF attempts these women with the syndrome undergo. It is also noteworthy that despite the fact that mothers with PCOS have an increased risk of gestational diabetes, fetal macrosomia was not mentioned [54].

Sir-Petermann et al. [51] showed a dramatic increase in concentrations of androgens (testosterone, androstenedione, and DHEA-S) in women with PCOS during pregnancy. Androgens are of partially placental and ovarian origin. These androgen concentrations do not result in fetal masculinisation which could be explained by the coexistence of increased levels of estradiol, progesterone, SHBG, and also because of placental aromatase that converts androstenedione to oestrone and 16-hydroxytestosterone to estriol. Carlsen et al. [10] showed that although intrauterine metformin exposure seems to result in elevated SHBG levels in neonates, maternal and neonatal androgen and estrogen levels remain unaffected.

The study of pregnancy outcomes associated with the use of metformin among women diagnosed with PCOS is also important [28]. The use of metformin has been shown to reduce spontaneous abortion rates [28, 43, 56] through hyperinsulinemia control [57, 58] and/or reduction of serum androgens levels [46]. Jakudowicz et al. studied the correlation of miscarriage rates in patients with PCOS treated or not with metformin during their pregnancy. Metformin was used in 65 pregnant women with PCOS, while 31 of the control group did not use it. The abortion rate was 8.8 and 41.9%, respectively [46]. Nawaz et al. [9] showed that continuation of metformin during pregnancy significantly reduces from 49.9 to 12.5% early pregnancy loss in obese women with PCOS. In another study the incidence of spontaneous abortion was 17% in 72 women—of whom 62% had a spontaneous abortion experience and whose conception was completed with metformin use either throughout pregnancy or for the first 12 weeks of gestation [49]. Nevertheless, an increased risk of spontaneous abortion up to 35% in women with PCOS has been referred despite metformin use [59], and it has also been referred that metformin use is associated with an increased risk of preeclampsia in women with PCOS [49]. Palomba et al. [14] in a recent metaanalysis showed that preconceptional use of metformin has no effect on the abortion risk of women with PCOS. When the side effects of metformin in the infant have been studied, no teratogenesis, weight, and height problems or growth and psychosocial developmental defects in their first 18 months of life were mentioned [25, 49]. Bolton et al. [16] did not find any difference in major congenital malformation and/or need of neonatal intensive care between women with PCOS who continued or discontinued metformin use in the first trimester of pregnancy. Although metformin is excreted in breast milk, its concentration is minimal [60–62]. However, the puerperants are recommended to receive metformin for 2–3 h after breastfeeding [63, 64].

Although, Siasiakos et al. [17] commented that there is a plausible mechanism for PCOS to cause adverse pregnancy outcome relating to the action of insulin, the studies preceding their review had not been rigorous enough to prove it, some new studies since, notably by Palomba et al. [8, 11], add evidence of better quality to the possible but never proven direct (not due to obesity or iatrogenic) link between PCOS and outcomes. For example, Palomba et al. [8] concluded that uterine artery Doppler indices such pulsatility index (PI) and bilateral notch at first and mid-second trimester are commonly altered in pregnant women with PCOS than in controls showing predictive adverse perinatal outcomes. Moreover, Palomba et al. [11] showed that there is higher relative risk for adverse obstetric or neonatal outcomes in patients with the full-blown and non-PCO phenotypes than in those with the non-hyperandrogenic and ovulatory phenotypes.

It should be noted, however, that most of the above-mentioned studies are retrospective having different criteria for the women included. For example, Toulis et al. [15] mentioned the significant statistical heterogeneity among different studies which study the risk of gestational diabetes mellitus in women with PCOS. For this reason, they concluded that a correlation between gestational diabetes mellitus and PCOS is questionable [15]. Some use only ultrasound findings for the diagnosis of PCOS and others hormonal or clinical findings. Furthermore, BMI and the age of the women were conflicting parameters. A randomised controlled study including three groups of women (women with PCOS according to Rotterdam criteria treated with metformin, untreated women with PCOS and normal pregnant women) could offer useful results of the effectiveness of metformin on perinatal outcome in pregnant women with PCOS.

Conclusion

The PCOS has an increased risk for maternal complications such as gestational diabetes, gestational hypertension, preeclampsia, and preterm labor. Especially, gestational diabetes is present independently of the obesity coexistence. The increased cesarean section rates of women with PCOS could be rather attributed to the obesity frequently seen in women with this syndrome. Perinatal morbidity and mortality are increased mainly due to the high incidence of maternal complications in women with PCOS. Metformin use seems to reduce the risk of miscarriage and diabetes mellitus. However, further studies are necessary in this field.