European Journal of Obstetrics & Gynecology and Reproductive Biology 181 (2014) 195–199
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Gestational diabetes mellitus risk factors in women with polycystic ovary syndrome (PCOS) Mahnaz Ashrafi a,b, Fatemeh Sheikhan a, Arezoo Arabipoor a, Roya Hosseini a, Fereshteh Nourbakhsh b,*, Zahra Zolfaghari a a
Department of Endocrinology and Female Infertility at Reproductive Biomedicine Research Center, Royan Institute for Reproductive Biomedicine, ACECR, Tehran, Iran Department of Obstetrics and Gynecology, School of Medicine, Tehran University of Medical Science, Tehran, Iran
b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 May 2013 Received in revised form 5 March 2014 Accepted 30 July 2014
Objectives: To compare the incidence of gestational diabetes mellitus (GDM) in Iranian infertile women with polycystic ovary syndrome (PCOS) and women without PCOS after pregnancies resulting from either assisted reproductive technology (ART) or spontaneous as well as to determine the risk factors of GDM in PCOS women. Study design: In a cross-sectional study, we evaluated medical records of 234 spontaneous pregnant women without PCOS in Akbarabadi Women’s Hospital affiliated to Tehran University of Medical Science, Tehran, Iran, along with 234 pregnant women with PCOS and 234 pregnant non-PCOS women with ART conception who were treated at Royan institute, Tehran, Iran, at the same period of time, 2012 to February 2013. Exclusion criteria were as following: maternal age 40, family history of diabetes in first-degree relatives, pre-pregnancy diabetes and history of gestational diabetes, history of stillbirth, recurrent miscarriage, birth weight baby 4 kg (macrosomia), parity >4, Cushing’s syndrome, congenital adrenal hyperplasia and overt hypothyroidism. The GDM diagnosis was according to American Diabetes Association (ADA) criteria. Incidence and the risk factors for GDM were evaluated. Results: The incidence rates of GDM were 44.4%, 29.9% and 7.3% for PCOS ART, non-PCOS ART and nonPCOS spontaneous pregnant women, respectively. Multivariable logistic regression was used for determining risk factors of GDM in PCOS women with adjusted odds ratios for age, parity and hypothyroidism, the results revealed the most important and significant predictors for development of GDM in PCOS women as follow: menstrual irregularity (OR = 4.2; 95% CI = 1.7–10.6), serum triglycerides level 150 mg/dL (OR = 1.9; 95% CI = 1.07–3.6) and pregestational metformin use (OR = 0.4; 95% CI = 0.2–0.7). Conclusions: Pregnant Iranian women with a history of infertility and PCOS are at increased risk for developing GDM. It is recommendable to perform screening test for GDM in PCOS women with ART treatment, irregular menses and high serum triglycerides level in the early stage of pregnancy. Pregestational use of metformin can be effective in reducing the occurrence of GDM. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Gestational diabetes mellitus Polycystic ovary syndrome Risk factors Infertility
Introduction Gestational diabetes mellitus (GDM), as a first recognition of carbohydrate intolerance, must be considered at onset or during the pregnancy. It approximately affects 0.2–8% of pregnancies and shows variations depending on ethnic/racial backgrounds and
* Corresponding author at: Department of Obstetrics and Gynecology, School of Medicine, Tehran University of Medical Science, Taleghani St., Ghods St., Enghelab Square, Tehran, 14176114411, Iran. Tel.: +98 21 88968349; fax: +98 21 88968349. E-mail address: [email protected] (F. Nourbakhsh). http://dx.doi.org/10.1016/j.ejogrb.2014.07.043 0301-2115/ß 2014 Elsevier Ireland Ltd. All rights reserved.
diagnostic criteria [1]. An imbalance between the capacity of pancreatic b-cells and the increased demands for insulin due to decreased insulin sensitivity during pregnancy is underlined for pathogenesis mechanisms of GDM [1–3]. Common risk factors for GDM include ethnicity/non-white race, obesity, older age and previous GDM history [4]. Other conditions with reduced insulin sensitivity and impaired b-cell function predisposing to glucose intolerance might also increase the risk of GDM [4]. Polycystic ovary syndrome (PCOS), as a complex genetic reproductive condition, is characterized by heterogeneous symptoms of clinical and/or chemical androgen excess, ovulatory dysfunction and polycystic ovaries [5]. The presence of insulin
196
M. Ashrafi et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 181 (2014) 195–199
resistance and impaired b-cell function are found in women with PCOS and in large number of young age patients exhibiting impaired glucose tolerance and non-insulin-dependent diabetes (NIDDM) [1]. Some studies reported an increase risk of glucose intolerance and type 2 diabetes in PCOS patients [4,6–8]. Some other studies reported that the risk of GDM is higher among PCOS versus non-PCOS women [4,9–13], but a number of studies did not find any association between the PCOS and GDM [14,15]. However, several studies presented an increased prevalence of PCOS in women with prior GDM, and also revealed a higher risk of developing GDM in PCOS women as compared to general population; in addition, a recent systematic review and metaanalysis reported that this finding is unreliable due to considerable heterogeneity among previous studies [16,17]. It was suggested that before making recommendation for pregnant women with PCOS regarding the risk of GDM, some properly designed studies should be conducted to assess the prevalence of GDM in PCOS women and to identify risk factors of gestational diabetes among them [10]. Considering the impact of race on outcomes, we conducted a cross-sectional study to evaluate the incidence and the risk factors for GDM in Iranian PCOS women and to identify the relationship between symptoms of this syndrome and gestational diabetes. Determining the specific risk factors of GDM in PCOS women is important for preventive intervention through screening tests at the early stages of pregnancy. Materials and methods This cross-sectional study was performed in reproductive biomedicine research center, Royan Institute, Tehran, Iran, from 2012 to February 2013. The inclusion criteria were women with singleton pregnancies, 13 weeks of gestation, as well as two-step screening for GDM and prenatal care performed either in Royan Institute or in Akbarabadi Women’s Hospital affiliated to Tehran University of Medical Science, Tehran, Iran, in the same period of time. Exclusion criteria were as following: maternal age 40, body mass index (BMI) 35 kg/m2, family history of diabetes in firstdegree relatives, pre-pregnancy diabetes, history of gestational diabetes, history of stillbirth, recurrent miscarriage, birth weight baby 4 kg (macrosomia), parity > 4, Cushing’s syndrome, congenital adrenal hyperplasia, untreated (overt) hypothyroidism and hyperprolactinemia. A total of 702 pregnant women recruited from either Royan Institute or Akbarabadi Women’s Hospital were divided into three groups according to mode of conception and PCOS diagnosis. Group 1 (n = 234) included the pregnant women with history of assisted reproductive technologies (ART) treatment and PCOS without other etiology of infertility. Group 2 (n = 234) included the pregnant women without PCOS and a history of infertility undergoing ART (in vitro fertilization and intracytoplasmic sperm injection cycle). Group 3 (n = 234) included the healthy pregnant women without PCOS and no history of infertility. Diagnosis of PCOS was performed based on presence of at least two of the Rotterdam criteria, including [18]: oligomenorrhea and or/amenorrhea; at least one characteristic of hyperandrogenism, such as hirsutism, acne, and sever hair loss; and polycystic ovaries detected by ultrasound. All the participants had fasting plasma glucose measurement in the first trimester of pregnancy. In our study settings, according to American Diabetes Association (ADA) criteria (2005) [19], pregnant women were screened at 24–28 weeks gestation using a 50 g 1-h oral glucose challenge test (OGCT). If the result of this screening test was (glucose 7.8 mmol/L or 140 mg/ dL), 1–2 weeks later, they had to take a 100 g 3-h oral glucose tolerance test (OGTT). Women were diagnosed with GDM after observing any of the following cut-off values based on the ADA criteria in result of 100 g OGTT: fasting 5.3 mmol/L (95 mg/dL); 1-h 10.0 mmol/L (180 mg/dL); 2-h 8.6 mmol/L (155 mg/dL);
and 3-h 7.8 mmol/L (140 mg/dL). The data concerning age, parity, history of diabetes during previous pregnancies and family history of diabetes were collected from the patients’ medical records. Normal value of serum thyrotropin hormone (TSH) was considered between 0.4 and 2.5 mIU/L. Hypothyroid women among study groups were treated with levothyroxine (LT4), pre-gestational and during pregnancy, these patients had TSH < 2.5 mIU/L before pregnancy. In the majority of women with polycystic ovary syndrome, metformin therapy was prescribed based on severity of PCOS and women’s BMI, three months before pregnancy. The study was approved by the Internal Review Board of Royan Institute. All patients signed a consent form on their initial visit giving permission to use their results without using their names in future studies. Statistical analysis Data were analyzed using Statistical Package for the Social Sciences (SPSS) version 16 (SPSS Inc. Chicago, IL). The distribution of quantitative variables was evaluated by Kolmogorov–Smirnov test, while normal or non-normal variables were analyzed by appropriate parametric or non-parametric tests. Categorical variables were compared using Chi-square test, and the comparisons of quantitative normal variables were performed by analysis of variance (ANOVA) test and student’s t-test among groups. Descriptive statistics are presented as mean standard deviation (SD), median/inter-quartile range and percentage. Logistic regression was applied to evaluate the risk factors for GDM in PCOS population and adjusted for potential confounding variables (age, parity and hypothyroidism). At least, 200 pregnant women were needed in each group to observe a difference of 10% in prevalence of GDM between women with PCOS and healthy women, with a power of 80%. Type I error was considered at size of 0.05. Results We evaluated medical records of 234 non-PCOS women with spontaneous pregnancies, 234 PCOS women and 234 non-PCOS women with ART pregnancies. We also compared the participants’ characteristics among three groups. As expected, age, menarche age, parity, irregular menses and hirsutism cases were significantly different among groups. No significant differences were found in terms of pre-pregnancy BMI and weight gain during pregnancy among groups. The incidence values of GDM were 44.4%, 29.9% and 7.3% for PCOS ART, non-PCOS ART and non-PCOS spontaneous pregnant women, respectively (p < 0.001). Also, the percentages of hypothyroidism cases under levothyroxine treatment were higher in PCOS (30.3%) and non-PCOS (24.4%) pregnant women treated by ART in comparison to non-PCOS spontaneous pregnant (1.3%) women (p < 0.001) (Table 1). Demographic and clinical characteristics of PCOS women with and without GDM were shown in Table 2. The two groups were significantly different in terms of age (p < 0.002) and irregular cycles (p < 0.01). Results showed that metformin consumption was associated with a reduction in GDM prevalence, significantly (p = 0.01). Pre-pregnancy laboratory tests and hormonal characteristics of PCOS women with and without GDM were compared in Table 4. The results showed that the two groups were significantly different in terms of mean serum levels of triglyceride and cholesterol (p = 0.02 and p = 0.04, respectively). The mean of serum free testosterone level is also close to significant (p = 0.09), which is more likely to reach a significant level by increasing the sample size. No statistically significant differences were observed in serum levels of luteinizing hormone (LH), follicle-stimulating hormone (FSH), thyroid-stimulating hormone (TSH), low-density lipoprotein (LDL),
M. Ashrafi et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 181 (2014) 195–199
197
Table 1 Comparison of the participants’ characteristics among three groups. Variables
PCOS pregnant ART group (n = 234)
Non-PCOS pregnant ART group (n = 234)
Non-PCOS pregnant spontaneous group (n = 234)
p-Value
Age (years) Menarche age (year) Irregular mense, n (%) Hirsutism score > 8a Pre-pregnancy BMI (kg/m2) Weight gain in pregnancy Parity, n (mean) Hypothyroidism cases under levothyroxine treatment (euthyroid subjects), n (%) Incidence of GDM, n (%)
29.6 3.9 13.3 1.5 198 (84.6) 128 (54.7) 26.1 3.4 11.4 2.4 1.3 0.6 71 (30.3)
30.7 4.7 13.4 2.0 26 (11.1) 24 (10.3) 25.5 4.2 11.3 2.3 1.4 0.8 57 (24.4)
26.4 5.5 12.4 1.2 8 (3.4) 9 (3.8) 25.7 3.8 11.3 2.6 2.0 1.2 3 (1.3)
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Gestational diabetes mellitus risk factors in women with polycystic ovary syndrome (PCOS) Mahnaz Ashrafi a,b, Fatemeh Sheikhan a, Arezoo Arabipoor a, Roya Hosseini a, Fereshteh Nourbakhsh b,*, Zahra Zolfaghari a a
Department of Endocrinology and Female Infertility at Reproductive Biomedicine Research Center, Royan Institute for Reproductive Biomedicine, ACECR, Tehran, Iran Department of Obstetrics and Gynecology, School of Medicine, Tehran University of Medical Science, Tehran, Iran
b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 May 2013 Received in revised form 5 March 2014 Accepted 30 July 2014
Objectives: To compare the incidence of gestational diabetes mellitus (GDM) in Iranian infertile women with polycystic ovary syndrome (PCOS) and women without PCOS after pregnancies resulting from either assisted reproductive technology (ART) or spontaneous as well as to determine the risk factors of GDM in PCOS women. Study design: In a cross-sectional study, we evaluated medical records of 234 spontaneous pregnant women without PCOS in Akbarabadi Women’s Hospital affiliated to Tehran University of Medical Science, Tehran, Iran, along with 234 pregnant women with PCOS and 234 pregnant non-PCOS women with ART conception who were treated at Royan institute, Tehran, Iran, at the same period of time, 2012 to February 2013. Exclusion criteria were as following: maternal age 40, family history of diabetes in first-degree relatives, pre-pregnancy diabetes and history of gestational diabetes, history of stillbirth, recurrent miscarriage, birth weight baby 4 kg (macrosomia), parity >4, Cushing’s syndrome, congenital adrenal hyperplasia and overt hypothyroidism. The GDM diagnosis was according to American Diabetes Association (ADA) criteria. Incidence and the risk factors for GDM were evaluated. Results: The incidence rates of GDM were 44.4%, 29.9% and 7.3% for PCOS ART, non-PCOS ART and nonPCOS spontaneous pregnant women, respectively. Multivariable logistic regression was used for determining risk factors of GDM in PCOS women with adjusted odds ratios for age, parity and hypothyroidism, the results revealed the most important and significant predictors for development of GDM in PCOS women as follow: menstrual irregularity (OR = 4.2; 95% CI = 1.7–10.6), serum triglycerides level 150 mg/dL (OR = 1.9; 95% CI = 1.07–3.6) and pregestational metformin use (OR = 0.4; 95% CI = 0.2–0.7). Conclusions: Pregnant Iranian women with a history of infertility and PCOS are at increased risk for developing GDM. It is recommendable to perform screening test for GDM in PCOS women with ART treatment, irregular menses and high serum triglycerides level in the early stage of pregnancy. Pregestational use of metformin can be effective in reducing the occurrence of GDM. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Gestational diabetes mellitus Polycystic ovary syndrome Risk factors Infertility
Introduction Gestational diabetes mellitus (GDM), as a first recognition of carbohydrate intolerance, must be considered at onset or during the pregnancy. It approximately affects 0.2–8% of pregnancies and shows variations depending on ethnic/racial backgrounds and
* Corresponding author at: Department of Obstetrics and Gynecology, School of Medicine, Tehran University of Medical Science, Taleghani St., Ghods St., Enghelab Square, Tehran, 14176114411, Iran. Tel.: +98 21 88968349; fax: +98 21 88968349. E-mail address: [email protected] (F. Nourbakhsh). http://dx.doi.org/10.1016/j.ejogrb.2014.07.043 0301-2115/ß 2014 Elsevier Ireland Ltd. All rights reserved.
diagnostic criteria [1]. An imbalance between the capacity of pancreatic b-cells and the increased demands for insulin due to decreased insulin sensitivity during pregnancy is underlined for pathogenesis mechanisms of GDM [1–3]. Common risk factors for GDM include ethnicity/non-white race, obesity, older age and previous GDM history [4]. Other conditions with reduced insulin sensitivity and impaired b-cell function predisposing to glucose intolerance might also increase the risk of GDM [4]. Polycystic ovary syndrome (PCOS), as a complex genetic reproductive condition, is characterized by heterogeneous symptoms of clinical and/or chemical androgen excess, ovulatory dysfunction and polycystic ovaries [5]. The presence of insulin
196
M. Ashrafi et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 181 (2014) 195–199
resistance and impaired b-cell function are found in women with PCOS and in large number of young age patients exhibiting impaired glucose tolerance and non-insulin-dependent diabetes (NIDDM) [1]. Some studies reported an increase risk of glucose intolerance and type 2 diabetes in PCOS patients [4,6–8]. Some other studies reported that the risk of GDM is higher among PCOS versus non-PCOS women [4,9–13], but a number of studies did not find any association between the PCOS and GDM [14,15]. However, several studies presented an increased prevalence of PCOS in women with prior GDM, and also revealed a higher risk of developing GDM in PCOS women as compared to general population; in addition, a recent systematic review and metaanalysis reported that this finding is unreliable due to considerable heterogeneity among previous studies [16,17]. It was suggested that before making recommendation for pregnant women with PCOS regarding the risk of GDM, some properly designed studies should be conducted to assess the prevalence of GDM in PCOS women and to identify risk factors of gestational diabetes among them [10]. Considering the impact of race on outcomes, we conducted a cross-sectional study to evaluate the incidence and the risk factors for GDM in Iranian PCOS women and to identify the relationship between symptoms of this syndrome and gestational diabetes. Determining the specific risk factors of GDM in PCOS women is important for preventive intervention through screening tests at the early stages of pregnancy. Materials and methods This cross-sectional study was performed in reproductive biomedicine research center, Royan Institute, Tehran, Iran, from 2012 to February 2013. The inclusion criteria were women with singleton pregnancies, 13 weeks of gestation, as well as two-step screening for GDM and prenatal care performed either in Royan Institute or in Akbarabadi Women’s Hospital affiliated to Tehran University of Medical Science, Tehran, Iran, in the same period of time. Exclusion criteria were as following: maternal age 40, body mass index (BMI) 35 kg/m2, family history of diabetes in firstdegree relatives, pre-pregnancy diabetes, history of gestational diabetes, history of stillbirth, recurrent miscarriage, birth weight baby 4 kg (macrosomia), parity > 4, Cushing’s syndrome, congenital adrenal hyperplasia, untreated (overt) hypothyroidism and hyperprolactinemia. A total of 702 pregnant women recruited from either Royan Institute or Akbarabadi Women’s Hospital were divided into three groups according to mode of conception and PCOS diagnosis. Group 1 (n = 234) included the pregnant women with history of assisted reproductive technologies (ART) treatment and PCOS without other etiology of infertility. Group 2 (n = 234) included the pregnant women without PCOS and a history of infertility undergoing ART (in vitro fertilization and intracytoplasmic sperm injection cycle). Group 3 (n = 234) included the healthy pregnant women without PCOS and no history of infertility. Diagnosis of PCOS was performed based on presence of at least two of the Rotterdam criteria, including [18]: oligomenorrhea and or/amenorrhea; at least one characteristic of hyperandrogenism, such as hirsutism, acne, and sever hair loss; and polycystic ovaries detected by ultrasound. All the participants had fasting plasma glucose measurement in the first trimester of pregnancy. In our study settings, according to American Diabetes Association (ADA) criteria (2005) [19], pregnant women were screened at 24–28 weeks gestation using a 50 g 1-h oral glucose challenge test (OGCT). If the result of this screening test was (glucose 7.8 mmol/L or 140 mg/ dL), 1–2 weeks later, they had to take a 100 g 3-h oral glucose tolerance test (OGTT). Women were diagnosed with GDM after observing any of the following cut-off values based on the ADA criteria in result of 100 g OGTT: fasting 5.3 mmol/L (95 mg/dL); 1-h 10.0 mmol/L (180 mg/dL); 2-h 8.6 mmol/L (155 mg/dL);
and 3-h 7.8 mmol/L (140 mg/dL). The data concerning age, parity, history of diabetes during previous pregnancies and family history of diabetes were collected from the patients’ medical records. Normal value of serum thyrotropin hormone (TSH) was considered between 0.4 and 2.5 mIU/L. Hypothyroid women among study groups were treated with levothyroxine (LT4), pre-gestational and during pregnancy, these patients had TSH < 2.5 mIU/L before pregnancy. In the majority of women with polycystic ovary syndrome, metformin therapy was prescribed based on severity of PCOS and women’s BMI, three months before pregnancy. The study was approved by the Internal Review Board of Royan Institute. All patients signed a consent form on their initial visit giving permission to use their results without using their names in future studies. Statistical analysis Data were analyzed using Statistical Package for the Social Sciences (SPSS) version 16 (SPSS Inc. Chicago, IL). The distribution of quantitative variables was evaluated by Kolmogorov–Smirnov test, while normal or non-normal variables were analyzed by appropriate parametric or non-parametric tests. Categorical variables were compared using Chi-square test, and the comparisons of quantitative normal variables were performed by analysis of variance (ANOVA) test and student’s t-test among groups. Descriptive statistics are presented as mean standard deviation (SD), median/inter-quartile range and percentage. Logistic regression was applied to evaluate the risk factors for GDM in PCOS population and adjusted for potential confounding variables (age, parity and hypothyroidism). At least, 200 pregnant women were needed in each group to observe a difference of 10% in prevalence of GDM between women with PCOS and healthy women, with a power of 80%. Type I error was considered at size of 0.05. Results We evaluated medical records of 234 non-PCOS women with spontaneous pregnancies, 234 PCOS women and 234 non-PCOS women with ART pregnancies. We also compared the participants’ characteristics among three groups. As expected, age, menarche age, parity, irregular menses and hirsutism cases were significantly different among groups. No significant differences were found in terms of pre-pregnancy BMI and weight gain during pregnancy among groups. The incidence values of GDM were 44.4%, 29.9% and 7.3% for PCOS ART, non-PCOS ART and non-PCOS spontaneous pregnant women, respectively (p < 0.001). Also, the percentages of hypothyroidism cases under levothyroxine treatment were higher in PCOS (30.3%) and non-PCOS (24.4%) pregnant women treated by ART in comparison to non-PCOS spontaneous pregnant (1.3%) women (p < 0.001) (Table 1). Demographic and clinical characteristics of PCOS women with and without GDM were shown in Table 2. The two groups were significantly different in terms of age (p < 0.002) and irregular cycles (p < 0.01). Results showed that metformin consumption was associated with a reduction in GDM prevalence, significantly (p = 0.01). Pre-pregnancy laboratory tests and hormonal characteristics of PCOS women with and without GDM were compared in Table 4. The results showed that the two groups were significantly different in terms of mean serum levels of triglyceride and cholesterol (p = 0.02 and p = 0.04, respectively). The mean of serum free testosterone level is also close to significant (p = 0.09), which is more likely to reach a significant level by increasing the sample size. No statistically significant differences were observed in serum levels of luteinizing hormone (LH), follicle-stimulating hormone (FSH), thyroid-stimulating hormone (TSH), low-density lipoprotein (LDL),
M. Ashrafi et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 181 (2014) 195–199
197
Table 1 Comparison of the participants’ characteristics among three groups. Variables
PCOS pregnant ART group (n = 234)
Non-PCOS pregnant ART group (n = 234)
Non-PCOS pregnant spontaneous group (n = 234)
p-Value
Age (years) Menarche age (year) Irregular mense, n (%) Hirsutism score > 8a Pre-pregnancy BMI (kg/m2) Weight gain in pregnancy Parity, n (mean) Hypothyroidism cases under levothyroxine treatment (euthyroid subjects), n (%) Incidence of GDM, n (%)
29.6 3.9 13.3 1.5 198 (84.6) 128 (54.7) 26.1 3.4 11.4 2.4 1.3 0.6 71 (30.3)
30.7 4.7 13.4 2.0 26 (11.1) 24 (10.3) 25.5 4.2 11.3 2.3 1.4 0.8 57 (24.4)
26.4 5.5 12.4 1.2 8 (3.4) 9 (3.8) 25.7 3.8 11.3 2.6 2.0 1.2 3 (1.3)
