Obstetric outcome in women with polycystic ovarian syndrome

M. Mikola, V. Hiilesmaa, M. Halttunen, L. Suhonen and A. Tiitinen,1

Department of Obstetrics and Gynaecology, Helsinki University Central Hospital, Helsinki, Finland


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Women with polycystic ovarian syndrome (PCOS) often have insulin resistance and hyperinsulinaemia and may therefore be at an increased risk for gestational diabetes mellitus (GDM). Hyperinsulinaemia may also be associated with pre-eclampsia. Information concerning outcome of pregnancies in PCOS women is scanty and somewhat controversial. Therefore, 99 pregnancies were retrospectively evaluated in women with PCOS and the findings were compared with an unselected control population. The average body mass index (BMI) in PCOS patients was greater than that in controls (25.6 versus 23.0) (P < 0.0001), and PCOS patients were more often nulliparous than controls (76 versus 42%) (P < 0.001). The multiple pregnancy rate was 9.1% in PCOS patients and 1.1% in controls [odds ratio (OR) 9.0; 95% confidence interval (CI) 3.5–23.3]. GDM developed in 20% of the PCOS patients and in 8.9% of the controls (P < 0.001). After logistic regression analysis, BMI >25 seemed to be the greatest predictor for GDM (adjusted OR 5.1; CI 3.2–8.3), while PCOS remained as another independent predictor (adjusted OR 1.9; CI 1.0–3.5). In contrast, PCOS was not a significant predictor for pre-eclampsia, which was merely associated with nulliparity. Premature delivery (16.1% in PCOS and 6.5% in controls) was explained to a large extent by multiple pregnancies and marginally by nulliparity and PCOS. In singleton pregnancies, there was no difference in birth weights, Apgar scores or perinatal morbidity of infants. In conclusion, PCOS slightly increases the risk for GDM, but does not have an important effect on the rate of premature delivery and pre-eclampsia.

Key words: gestational diabetes mellitus/polycystic ovary syndrome/pre-eclampsia/pregnancy


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Polycystic ovarian syndrome (PCOS) is characterized by menstrual irregularities, hirsutism, and obesity. Increased serum concentrations of androgens, LH and insulin are key features of its endocrine profile (Conway et al., 1990Go; Falcone et al., 1992Go; Anttila et al., 1993Go; Holte et al., 1994aGo, bGo; Lanzone et al., 1996Go). Anovulatory infertility is frequent; however, most PCOS patients conceive successfully following ovulation induction. Insulin resistance in non-pregnant and pregnant PCOS patients has been demonstrated in many reports (Lanzone et al., 1996Go, 1998; Holte et al., 1998Go; Paradisi et al., 1998Go).

The risks due to multiple pregnancies after ovulation induction have been evaluated in PCOS patients (Balen et al., 1994Go). There are, however, few reports concerning other obstetric complications, and the numbers of pregnancies per study have been small. A higher than expected risk of pre-eclampsia has been reported in some studies (Diamant et al., 1982Go; Gjonnaess, 1989Go; de Vries et al., 1998Go; Radon et al., 1999Go).

Since pregnancy induces insulin resistance (Knopp et al., 1981Go), PCOS patients may also be at risk for developing gestational diabetes mellitus (GDM). No increase in the prevalence of GDM was observed in one study (Wortsman et al., 1991Go), whereas polycystic ovaries were common among women with previous GDM in two other studies (Anttila et al., 1998Go; Holte et al., 1998Go). A recent study showed an increased risk for GDM in PCOS pregnancies (Radon et al., 1999Go). The scanty and partly conflicting information on PCOS pregnancies prompted this study to evaluate the obstetric outcome of PCOS pregnancies.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Between 1 January, 1988 and 31 December, 1998, 99 pregnancies in 80 PCOS patients were treated at the Department of Obstetrics and Gynaecology of Helsinki University Central Hospital. The medical records were retrospectively reviewed for each patient. Seventy-seven deliveries took place at our hospital, and 22 elsewhere. The obstetric outcome was assessed by hospital records and a questionnaire sent to the patients. The couples had had, on average, 3 years of infertility and 72% (71/99) of the pregnancies studied resulted from ovulation induction. The local ethics committee approved this study.

PCOS was diagnosed according to the following criteria: (i) ultrasonographic appearance of polycystic ovaries (Adams et al., 1986Go) and at least two of the following: (ii) serum LH/FSH ratio >2; (iii) elevated androgen concentrations [testosterone, free testosterone, androstenedione or dehydroepiandrosterone sulphate (DHEAS)]; or (iv) clinical picture of menstrual irregularities, hirsutism, or infertility from anovulation (Tiitinen et al., 1994Go). Women with body mass index (BMI) of <25 kg/m2 prior to pregnancy were classified as normal weight and those with BMI >25 as obese.

Gestational diabetes was diagnosed by using the 75 g, 2 h oral glucose tolerance test (OGTT). Patients were classified as having gestational diabetes mellitus (GDM) if one of the following reached or exceeded the 97.5 percentile: fasting venous blood glucose concentration (4.5 mmol/l), or 1 h (9.1 mmol/l), or 2 h (7.9 mmol/l). For capillary blood and venous plasma the threshold values were 4.8 mmol/l, 10.0 mmol/l, and 8.7 mmol/l respectively (Teramo et al., 1993Go). The indications for routine OGTT during the second trimester were: (i) glucose in any of the monthly urine samples, (ii) GDM in previous pregnancy, (iii) diabetes in the immediate family, (iv) obesity (BMI >25), or (v) fetal macrosomy (>2 SD) in index or previous pregnancies.

Pre-eclampsia was diagnosed as blood pressure with systole >140 mmHg or diastole >90 mmHg during the third trimester of pregnancy, and accompanying proteinuria (>0.3 g/24 h). Deliveries at <37 weeks of gestation were defined as premature.

To obtain a representative sample of the background population, 737 pregnancies in 712 women were used as controls. The controls were residents of Kerava, a town within the catchment area of our hospital, who were identified in 1993–94 as they attended routine ultrasound screening at 16–18 weeks gestation. About 95% of pregnant women in the area underwent this examination. The only exclusion criteria for controls were miscarriage or abortion before 22 weeks of gestation.

Continuous variables were compared using the Student's t-test or the Mann–Whitney U-test, and categoric data using the {chi}2 test with Yates' correction, or Fisher's exact probability test. Multiple logistic regression analyses were performed with pre-eclampsia, GDM or premature delivery as dependent variables. Odds ratios (OR) with 95% confidence intervals (CI) were estimated for the five independent predictor variables (PCOS, BMI, parity, maternal age, and number of fetuses). P values of < 0.05 were considered statistically significant. The calculations were done using NCSS 2000 (NCSS, Kaysville, UT, USA).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Women with PCOS were 7.2 kg heavier and 1 year older than controls. Seventy-six per cent of PCOS women were nulliparous compared with 42% of controls (Table IGo). All our patients had either clinical or laboratory signs of hyperandrogenism. A total of 44 pregnancies were induced with clomiphene citrate, 25 with gonadotrophin stimulation, and one with IVF and embryo transfer.


View this table:
[in this window]
[in a new window]
 
Table I. General characteristics of patients
 
Nine (9%) multiple pregnancies occurred in the PCOS group: seven sets of twins and two sets of triplets. In the control group, there were eight multiple pregnancies: seven sets of twins and one set of triplets. The odds ratio for multiple pregnancy in women with PCOS was 9.0 (95% CI 3.5–23).

Among singletons, there were no differences in mean birth weight, length, head circumference, and Apgar scores (Table IIGo). Birth weights adjusted for gestational weeks and sex (SD scores) conformed to those of the Finnish general population. Infants weighing > 4500 g comprised 3.6% and 3.5% of the PCOS and control groups respectively.


View this table:
[in this window]
[in a new window]
 
Table II. General characteristics of offspring in singleton pregnancies
 
In multiple pregnancies, the average birth weight in the PCOS group was 2078 g (range 1515–2720 g) as compared with 2536 g (range 1500–3585 g) in the control group.

Caesarean section rates for term singleton pregnancies were 28% among the PCOS group compared with 17% for the controls (OR 2.1; CI 1.4–3.4; P = 0.0008). The breakdown of the indications for Caesarean sections (abnormal presentation, prolonged labour, fetal asphyxia and miscellaneous) was similar among the PCOS and control groups. For premature births, the Caesarean section rates were 50% and 24% respectively (not significant).

There were two perinatal deaths (2.0%) in the PCOS group: a delivery at 24 weeks due to chorionamnitis, and one Potter's syndrome. In the control group, there were three perinatal deaths (0.4%) comprising one trisomy 18, one left heart hypoplasia, and one placental ablation at 22 weeks. The difference in perinatal mortality rates was not statistically significant.

A total of 32% of the women with PCOS and 35% of the controls had OGTT during pregnancy (not significant); 20% of the women with PCOS and 9.0% of the controls developed GDM (P < 0.001). After logistic regression, BMI >25 emerged as the most important predictor, but the effect of PCOS also remained statistically significant (Table IIIGo). Four type I insulin-dependent diabetics among controls and one type II diabetic among PCOS patients were excluded in this analysis. Three out of 20 women with GDM (15%) in the PCOS group required insulin treatment compared with five of 66 (8%) in the control group. In singleton pregnancies PCOS showed an OR of 1.6 (95% CI 0.9–3.0) after adjustment of BMI only. Among obese women (BMI >25) GDM developed in 29% of the PCOS patients and in 22% of the controls (P = NS). In lean subjects (BMI <=25) the figures were 20% and 9% respectively (P = NS).


View this table:
[in this window]
[in a new window]
 
Table III. Risk factors for gestational diabetes mellitus as assessed by logistic regression
 
Four PCOS (4%) and 14 control pregnancies (2%) were complicated by pre-eclampsia (RR 2.2; 95% CI 0.7–6.7). After multiple logistic regression analyses, nulliparity appeared to be the only significant risk factor for pre-eclampsia, while BMI >25, age, multiple pregnancy, or PCOS had no predictive value (Table IVGo).


View this table:
[in this window]
[in a new window]
 
Table IV. Risk factors for pre-eclampsia as assessed by logistic regression
 
The rate for prematurity was 16.1% in PCOS patients and 6.5% in controls (P < 0.001). In singleton pregnancies, 10 PCOS (11%) and 45 control (6%) women delivered before the 37th gestational week (OR 1.8; CI 0.9–3.3). The greatest predictor for premature delivery was multiple pregnancy (adjusted OR 11.1; CI 3.9–31.2), while nulliparity had a marginal effect (adjusted OR 1.8; CI 1.0–3.2). PCOS had an effect which, however, did not quite reach statistically significant level after adjustment for multiple pregnancy and nulliparity (Table VGo).


View this table:
[in this window]
[in a new window]
 
Table V. Risk factors for premature delivery as assessed by logistic regression
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A slightly increased risk of GDM in PCOS pregnancies was found. The higher incidence of GDM in PCOS was largely explained by obesity and, to a small extent of borderline statistical significance, independently by PCOS itself. The same criteria for performing an OGTT were applied regardless of whether or not the woman had PCOS and, indeed, OGTT was not performed in our PCOS patients more frequently.

Similar findings have earlier been reported by others (Gjonnaess, 1989Go), but the results were explained by obesity rather than PCOS. An increased incidence of GDM has recently been reported in a smaller study (Radon et al., 1999Go). In this retrospective study, 22 PCOS women were compared with 66 control women matched for age and weight, but not for parity. Nine of 22 women with PCOS developed GDM during pregnancy (OR 22.2, 95% CI 3.8–170.0). In contrast, in another study, no increased risk for GDM or fetal macrosomia in PCOS patients was found (Wortsman et al., 1991Go).

In women with previous GDM, the prevalence of polycystic ovaries (PCO) was increased when studied by ultrasonography some years after the index pregnancy (Anttila et al., 1998Go; Holte et al., 1998Go). Some of these women also had features of the clinical syndrome. The ultrasonographic findings were suggestive of PCO in 45% (Anttila et al., 1998Go) and 41% (Holte et al., 1998Go) of women with previous GDM. This prevalence is twice as large as that reported for premenopausal women overall. These results strongly suggest an association between PCO and GDM. When the endocrine profile of the patients with previous GDM was studied, those with PCOS had evidence of insulin resistance (Holte et al., 1998Go). Due to the retrospective character of the current study, it was not possible to study the correlation between the prepregnancy insulin resistance and the occurrence of GDM, since insulin concentrations were measured in only a few of the PCOS patients.

At least one third of PCOS patients are overweight (Conway et al., 1989Go; Franks, 1989Go), and in the current study 52% had prepregnancy BMI >25. Obesity is associated with hyperinsulinaemia and insulin resistance; however, similar abnormalities in carbohydrate metabolism have also been demonstrated in lean PCOS patients (Dunaif et al., 1989Go; Lanzone et al., 1990Go; Falcone et al., 1992Go; Anttila et al., 1993Go). A possible explanation for the increased risk of GDM in women with PCOS is this altered insulin metabolism, which is partly independent of body weight. Early alteration of insulin sensitivity and the compensatory insulin hypersecretion constitute specific risk factors in PCOS patients for the development of abnormalities of glucose tolerance (Paradisi et al., 1998Go).

There are reports on an increased risk for pre-eclampsia in PCOS (Gjonnaess, 1989Go; Radon et al., 1999Go), but parity has not been accounted for. Thus, the confounding effect of nulliparity cannot be excluded. Another study (de Vries et al., 1998Go), using matching for age and parity, found an increased incidence of pre-eclampsia in PCOS. However, the overall incidence of pregnancy-induced hypertension was not significantly different between PCOS and control groups (26 and 25% respectively). In contrast, the findings presented here do not suggest that PCOS alone increases the risk of pre-eclampsia, which was merely associated with nulliparity and multiple pregnancy. On the other hand, women with prior pre-eclampsia have been shown to be hyperinsulinaemic and hyperandrogenic later in life (Laivuori et al., 1996Go, 1998Go), which also suggests a possible association between these two disorders.

Ovulation induction and the related increase in multiple pregnancies obviously explain premature delivery in PCOS pregnancies. The multiple pregnancy rate was 9% across all PCOS pregnancies and 12.5% in those treated with ovulation induction, which agrees with previous reports (Balen et al., 1994Go). Premature deliveries in singleton pregnancies did not differ markedly between PCOS patients and controls; however, due to small numbers, a three-fold increase among PCOS pregnancies would still be compatible with the results of this study.

In this series, the rate of Caesarean sections in term singleton pregnancies was significantly higher (28%) in the PCOS group compared with 17% in controls. This difference, however, could not be explained by corresponding increase in maternal or fetal complications. No differences were observed in the birth weights or Apgar scores of the infants. GDM is often associated with macrosomia. No evidence of this was found in the present study, but we believe it is because of preventive measures, including careful monitoring and dietary advice during the pregnancy. The similar Apgar scores, birth weights, and perinatal mortality rates among the PCOS group and controls suggest that no markedly increased risk for perinatal morbidity or mortality exists in singleton PCOS pregnancies.

Nulliparity and multiple pregnancies explain the higher rates of pre-eclampsia and premature delivery. However, there is an increased risk for GDM in PCOS pregnancies, which is largely explained by obesity, but also partly by the underlying PCOS. Since GDM may be associated with neonatal morbidity, its screening during second-trimester PCOS pregnancies seems reasonable.


    Notes
 
1 To whom correspondence should be addressed at: Department of Obstetrics and Gynaecology, Helsinki University Central Hospital, PO Box 140, FIN-00029 HUS, Finland. E-mail: aila.tiitinen{at}hus.fi Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Adams, J., Polson, D.W. and Franks, S. (1986) Prevalence of polycystic ovaries in women with anovulation and idiopathic hirsutism. Br. Med. J., 293, 335–359.

Anttila, L., Koskinen, P., Jaatinen, T. et al. (1993) Insulin hypersecretion together with high luteinizing hormone concentration augments androgen secretion in oral glucose tolerance test in women with polycystic ovarian disease. Hum. Reprod., 8, 1179–1183.[Abstract]

Anttila, L., Karjala, K., Penttilä, T-A. et al. (1998) Polycystic ovaries in women with gestational diabetes. Obstet. Gynecol., 92, 13–16.[Abstract/Free Full Text]

Balen, A.H., Braat, D.D., West, C. et al. (1994) Cumulative conception and live birth rates after treatment of anovulatory infertility: safety and efficacy of ovulation induction in 200 patients. Hum. Reprod., 9, 1563–1570.[Abstract]

Conway, G.S., Honour, J.W. and Jacobs, H.S. (1989) Heterogeneity of the polycystic ovary syndrome: clinical, endocrine and ultrasound features in 556 patients. Clin. Endocrinol., 30, 459–470.[ISI][Medline]

Conway, G.S., Jacobs, H.S., Holly, J.P.M. et al. (1990) Effects of luteinizing hormone, insulin, insulin-like growth factor-1 and insulin-like growth factor-1 binding protein-1 in the polycystic ovary syndrome. Clin. Endocrinol., 33, 593–603.[ISI][Medline]

de Vries, M.J., Dekker, G.A. and Schoemaker, J. (1998) Higher risk of pre-eclampsia in the polycystic ovary syndrome. A case control study. Eur. J. Obstet. Gynecol. Reprod. Biol., 76, 91–95.[ISI][Medline]

Diamant, Y.Z., Rimon, E. and Evron, S. (1982) High incidence of pre-eclamptic toxemia in patients with polycystic ovarian disease. Eur. J. Obstet. Gynec. Reprod. Biol., 14, 199–204.[ISI]

Dunaif, A., Segal, K.R., Futterweit, W. et al. (1989) Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes, 38, 1165–1174.[Abstract]

Falcone, T., Little, A.B. and Morris, D. (1992) Impaired glucose effectiveness in patients with polycystic ovary syndrome. Hum. Reprod., 7, 922–925.[Abstract]

Franks, S. (1989) Polycystic ovary syndrome: a changing perspective. Clin. Endocrinol., 31, 87–120.[ISI][Medline]

Gjonnaess, H. (1989) The course and outcome of pregnancy after ovarian electrocautery in women with polycystic ovarian syndrome: the influence of body weight. Br. J. Obstet. Gynaecol., 96, 714–719.[ISI][Medline]

Holte, J., Bergh, T., Berne, C. et al. (1994a) Enhanced early insulin response to glucose in relation to insulin resistance in women with polycystic ovary syndrome and normal glucose tolerance. J. Clin. Endocrinol. Metab., 78, 1052–1058.[Abstract]

Holte, J., Bergh, T., Berne, C. et al. (1994b) Serum lipoprotein lipid profile in women with the polycystic ovary syndrome – relation to anthropometric, endocrine and metabolic variables. Clin. Endocrinol., 41, 463–471.[ISI][Medline]

Holte, J., Gennarelli, G., Wide, L. et al. (1998) High prevalence of polycystic ovaries and associated clinical, endocrine, and metabolic features in women with previous gestational diabetes mellitus. J. Clin. Endocrinol. Metab., 83, 1143–1150.[Abstract/Free Full Text]

Knopp, R.H., Montes, A., Childs, M. et al. (1981) Metabolic adjustments in normal and diabetic pregnancy. Clin. Obstet. Gynecol., 24, 21–49.[Medline]

Laivuori, H., Kaaja, R., Rutanen, E-M. et al. (1998) Evidence of high circulating testosterone in women with prior pre-eclampsia. J. Clin. Endocrinol. Metab., 83, 344–347.[Abstract/Free Full Text]

Laivuori, H., Tikkanen, M.J. and Ylikorkala, O. (1996) Hyperinsulinemia 17 years after pre-eclamptic first pregnancy. J. Clin. Endocrinol. Metab., 81, 2908–2911.[Abstract]

Lanzone, A., Fulghesu, A.M., Andreani, C.L. et al. (1990) Insulin secretion in polycystic ovarian disease: effect of ovarian suppression by GnRH-agonist. Hum. Reprod., 5, 143–149.[Abstract]

Lanzone, A., Fulghesu, A.M., Cuccinelli, F. et al. (1996) Preconceptional and gestational evaluation of insulin secretion in patients with polycystic ovary syndrome. Hum. Reprod., 11, 2382–2386.[Abstract]

Paradisi, G., Fulghesu, A.M., Ferrazani, S. et al. (1998) Endocrino-metabolic features in women with polycystic ovary syndrome during pregnancy. Hum. Reprod., 13, 542–546.[Abstract]

Radon, P.A., McMahon, M.J. and Meyer, W.R. (1999) Impaired glucose tolerance in pregnant women with polycystic ovary syndrome. Obstet. Gynecol., 94, 194–197.[Abstract/Free Full Text]

Teramo, K., Haukkamaa, M., Leinonen, P. et al. (1993) Gestaatiodiabetes, Suomen Diabetesliiton työryhmän suositus (recommendations on gestational diabetes by the committee of the Finnish diabetes union). Suomen Lääkärilehti, 26, 2451–2454.

Tiitinen, A., Simberg, N., Stenman, U-H. et al. (1994) Estrogen replacement does not potentiate gonadotropin-releasing hormone agonist-induced androgen suppression in treatment of hirsutism. J. Clin. Endocrinol. Metab., 79, 447–451.[Abstract]

Wortsman, J., de Angeles, S., Futterweit, W. et al. (1991) Gestational diabetes and neonatal macrosomia in the polycystic ovary syndrome. J. Reprod. Med., 36, 659–661.[ISI][Medline]

Submitted on June 1, 2000; accepted on October 13, 2000.