Clinical Endocrinology (2014) 81, 93–99

doi: 10.1111/cen.12406

ORIGINAL ARTICLE

Metabolic aspects of different phenotypes of polycystic ovary syndrome: Iranian PCOS Prevalence Study Farhad Hosseinpanah*, Maryam Barzin*, Sorena Keihani†, Fahimeh Ramezani Tehrani† and Fereidoun Azizi‡ *Obesity Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, †Reproductive Endocrinology Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, and ‡Endocrine Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran

Summary Objective To compare insulin resistance (IR) and metabolic characteristics of different phenotypes of polycystic ovary syndrome (PCOS) and healthy controls in a representative sample of reproductive-aged Iranian women. Design Cross-sectional community-based study. Patients Of 915 participants, 136 met Rotterdam criteria for PCOS and were divided in four subgroups: ovulatory dysfunction with polycystic ovaries (Anovu + PCO), hyperandrogenism with PCO (HA + PCO), hyperandrogenism with ovulatory dysfunction (HA + Anovu) and full Rotterdam (HA + Anovu + PCO). Also 423 subjects served as normal controls, and the rest of participants were analysed as isolated HA, Anovu or PCO subgroups. Measurements Anthropometric measures, biochemical parameters and IR were compared between different groups. IR was defined using the homeostatic model assessment-IR (HOMAIR). Results Overall prevalence of IR in PCOS and normal women were 26
5% and 23
9%, respectively (P = 0
03). After age and BMI adjustment, there was no difference regarding metabolic characteristics between different PCOS phenotypes. Also metabolic characteristics, fasting insulin and HOMA-IR were comparable in PCOS subjects with hyperandrogenism and the nonhyperandrogenic phenotype (Anovu + PCO). In regression analysis, only overweight and obesity (BMI ≥ 25 kg/m2) had an independent association with IR (OR: 2
49, CI: 1
06–5
82, P = 0
03). Conclusions In a representative sample of an Iranian PCOS population, there is no significant difference between different PCOS phenotypes or between PCOS women and normal subjects regarding their metabolic characteristics. These findings suggest that overall metabolic screening may not always be

Correspondence: Fahimeh Ramezani Tehrani, Director, Professor, Reproductive Endocrinology Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, P.O. Box: 19395-476, Tehran, Iran. Tel.: +98 21 22432500; Fax: +98 21 22416264; E-mail: [email protected] © 2014 John Wiley & Sons Ltd

required for all women diagnosed with PCOS. Larger-scale population-based studies are needed to further assess these issues. (Received 1 November 2013; returned for revision 28 November 2013; finally revised 8 January 2014; accepted 8 January 2014)

Introduction Polycystic ovary syndrome (PCOS) is one of the most common endocrine disorders affecting about 1 in every 15 women worldwide.1 PCOS affects many organ systems and can result in several health complications including hirsutism, acne, menstrual dysfunction and unfavourable profiles for cardio-metabolic risk factors.1 While there is an impressive body of research showing higher prevalence of metabolic syndrome in patients with PCOS,2,3 some studies report a lack of or a weak association between PCOS and metabolic syndrome.4,5 Moreover, it has been shown that various phenotypes of PCOS differ regarding anthropometric, hormonal and metabolic indices.6–9 Two newer phenotypes of PCOS proposed by the Rotterdam criteria, specifically the anovulatory with polycystic ovaries (Anovu + PCO) phenotype, which were not included in the previous NIH definition of PCOS, merit some special consideration in this regard. Some studies suggest that lack of HA represents less severe forms of PCOS in terms of metabolic abnormalities. For instance, Zhang et al.7 showed that the nonhyperandrogenic PCOS phenotype (Anovu + PCO) might represent a phenotype of PCOS associated with a more favourable metabolic profile, compared with the other phenotypes. Goverde et al.10 reported that hyperandrogenic PCOS phenotypes are highly linked to the presence of metabolic syndrome and insulin resistance compared with nonhyperandrogenic phenotypes. However, some other studies report no significant metabolic differences between various PCOS phenotypes.11,12 In our previous study, we have reported a lack of association between PCOS and metabolic syndrome.4 Given the aforementioned inconsistency between different studies, we aimed to compare insulin resistance (IR) and metabolic characteristics of different PCOS phenotypes and healthy controls in a representative sample of reproductive-aged Iranian women. 93

94 F. Hosseinpanah et al.

Methods Study subjects The subjects of this study were selected from the Iranian PCOS Prevalence Study, a population-based cross-sectional study of 1126 women, aged 18–45 years, undertaken between 2009 and 2010.13 The stratified, multistage probability cluster sampling, with a probability in proportion to size procedure was used for recruitments of study subjects. Details of the inclusion and exclusion criteria have been published elsewhere.13 In brief, menopausal women (n = 37), and pregnant women at the time of the evaluation (n = 53) were excluded, and eligible women were invited for a comprehensive interview and blood pressure, anthropometric, hormonal and metabolic measurements. These data were completed for all except 97 women who did not visit clinics. The hormonal and metabolic profiles of 10 women were not available. We further excluded those women who had hyperprolactinaemia (n = 4) and women with thyroid dysfunction (n = 10). Finally, the 915 remaining participants constituted our five study groups as follows: PCOS by Rotterdam (Rott) criteria (n = 136), women with only androgen excess (n = 205), women with only ovulatory dysfunction (n = 77), women with only polycystic ovaries (PCO) (n = 74) and healthy controls (n = 423). Women with PCOS by Rotterdam criteria (n = 136) were further divided in four subgroups: ovulatory dysfunction with polycystic ovaries (Anovu + PCO) (n = 27), HA with PCO (HA + PCO) (n = 43), HA with ovulatory dysfunction (HA + Anovu) (n = 54), and having all the three components as full Rotterdam (full Rott, that is, HA + Anovu + PCO) (n = 12). Measurements All participants underwent clinical examinations, where trained staff measured body weight, height, waist, hip circumferences and blood pressure. Height and weight were measured with subjects in light clothes and without shoes, using standard apparatus. Weight was measured to the nearest 0
1 kg on a calibrated beam scale. Height and waist circumference (WC) were measured to the nearest 0
5 cm with a measuring tape. Waist was measured midway between the lower rib margin and the iliaccrest at the end of a gentle expiration. BMI was calculated as weight in kilograms, divided by the height in metres squared (kg/m2). All of the study subjects were invited for transvaginal (n = 760) or transabdominal (n = 155) ultrasound of the ovaries, performed using the 3
5-MHz transabdominal and 5-MHz transvaginal transducer by an experienced sonographer and a single sonographer assessed all scans. Ultrasound was performed on the same day as the blood samples were collected. A blood sample was taken after 12-h overnight fasting for biochemical measurements. Samples were taken from each subject on the second or third day of their spontaneous or progesterone-induced menstrual cycles. Blood samples were collected in EDTA-treated test tubes. Plasma was separated in a refrigerated centrifuge at 1560 g for 10 min, and sera were stored at 80 °C

until tested. Fasting plasma glucose (FPG) was measured using an enzymatic colorimetric method with glucose oxidase. Lipid profiles including total cholesterol (TC), triglycerides (TG) and high-density lipoprotein cholesterol (HDL-C) were measured by commercial assay kits (Pars Azmoon Inc., Tehran, Iran). TC and TG were assayed using enzymatic colorimetric tests with cholesterol esterase and cholesterol oxidase, and glycerol phosphate oxidase, respectively. HDL-C was measured after precipitation of the apolipoprotein-B containing lipoproteins with phosphotungstic acid. All samples were analysed when internal quality control met the acceptable criteria. Insulin was assessed by Immuno Enzyme Metric Assay (IEMA) (Mercodia, Uppsala, Sweden). Dehydroepiandrosterone sulphate (DHEAS) and total testosterone (TT) were measured by enzyme immune assay (Diagnostic Biochem Canada Co. Ontario, Canada). Sex-hormone-binding globulin (SHBG) was measured by IEMA (Diagnostic Biochem Canada Co.). All ELISA tests were carried out using Sunrise ELISA reader (Tecan Co. Salzburg, Austria). Luteinizing hormone (LH) and follicle-stimulating hormone (FSH) were measured by Immunoradimetric Assay [(Izotop, Budapest, Hungary) using gamma counter WallacWizard, Turku, Finland]. The free androgen index (FAI) was calculated using the formula [TT (nm/l) 9 100/SHBG (nm/l)]. The intra- and interassay coefficients of variation were 2
5% and 3
2% for TC, 2
2% and 2
3% for TG, 5
6% and 6
6% for TT, 5
1% and 6
8% for DHEAS, 1
2% and 5
7% for SHBG, 3% and 5
8% for LH and 3
5% and 4
0% for FSH, respectively. Definitions We defined PCOS using the Rotterdam Criteria14 by the presence of two or more of the following: (I) oligo- and/or anovulation (Anovu), menstrual cycles at no 95 cm (%) High BP (%) HDL-C < 50 mg/dl (%) Triglycerides ≥ 150 mg/dl (%) FBS ≥ 100 mg/dl (%) IR† (%)

characteristics

between

PCOS

PCOS with HA* n = 109

PCO-Anovu n = 27

P value

31
9 26
9 84
9 107
3 68
5 83
9 180
3

34
6 26
7 86
8 107
0 66
8 89
2 175
2

(7
1) (4
6) (12
0) (9
8) (9
3) (23
1) (32
1)

0
04 0
29 0
24 0
53 0
31 0
26 0
17

(11
9) (86–158) (4
2–10
0) (0
8–2
8)

0
64 0
15 0
36 0
11 – – –

44
0 124 8
2 1
7 26
9 4
6 70
4

(7
7) (6
1) (13
6) (13
9) (11
4) (14
7) (36
1) (12. 3) (85–180) (5
6–11
5) (1
1–2
5)

45
3 115 8
3 1
7 33
3 0
0 51
9

32
4

33
3

–

14
8

18
5

–

25
9

29
6

–

Values are given as mean (SD) if otherwise median (25–75 interquartile). PCOS, polycystic ovary syndrome; Anovu, Anovulation; HA, Hyperandrogenism; BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; HOMA-IR, homeostasis model assessment–insulin resistance; IR, insulin resistance. *Clinical and/or biochemical HA. †Prevalence rate of IR based on cut-off of 2
30 mol 9 lU/l2 for HOMA-IR.

Table 3. Logistic regression analysis of predictive development of IR* in PCOS women (n = 136)

Univariable analysis Age (years) Overweight & obesity (Ref, BMI < 25 kg/m2) Abdominal obesity (Ref, WC