Translate this page into:
Lipid variations in different polycystic ovary syndrome phenotypes: A systematic review & meta-analysis
For correspondence: Dr Prabhaker Mishra, Department of Biostatistics and Health Informatics, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow 226 014, Uttar Pradesh, India e-mail: drpmishrasgpgi@gmail.com
-
Received: ,
Accepted: ,
Abstract
Background & objectives
Polycystic Ovary Syndrome (PCOS) is an endocrine disorder affecting reproductive-age women worldwide. Lipid abnormalities, such as elevated low-density lipoprotein (LDL) and triglyceride (TG) levels and reduced high-density lipoprotein (HDL) levels, are commonly observed in women with PCOS, increasing their risk of cardiovascular disease (CVD). Therefore, this study aims to quantify the magnitude and pattern of lipid levels (total cholesterol, HDL, LDL, and TG) in women with different phenotypes of PCOS versus control women.
Methods
Worldwide published observational (cross-sectional, case-control, and cohort) studies between January 2010 and December 2024 were systematically searched and assessed using electronic databases, such as PubMed, Google Scholar, Science Direct, and Web of Science-Science Citation Index, where women suffering from different PCOS phenotypes were compared with non-PCOS controls. The association between lipid levels and PCOS was estimated by the mean difference (MD) with a 95% confidence interval (CI).
Results
The studies included 3655 PCOS patients (phenotype A 1907, phenotype B 474, phenotype C 764, phenotype D 510) and 1824 control participants. Women with the complete phenotype polycystic ovarian morphology plus hyperandrogenism plus ovulatory dysfunction (PCO+HA+O) had increased levels of total cholesterol, LDL cholesterol, and TGs compared to women with other PCOS phenotypes. Total cholesterol was 12.69 mg/dl [95% confidence interval (CI): 8.25, 17.13] in phenotype A. TG levels exhibited the greatest MD in phenotype A and the smallest in phenotype C when compared to control subjects.
Interpretation & conclusions
The study found significant differences in lipid levels among different PCOS phenotypes compared to control women, highlighting the significance of recognising these differences for cardiovascular risk management.
Keywords
Dyslipidaemia
hyperandrogenism
PCOS
phenotypes
Polycystic Ovary Syndrome (PCOS) is one of the most prevalent metabolic-hyperandrogenic-ovulatory disorders, affecting approximately 6 to 12 per cent of women of reproductive age worldwide1-3. Based on the variety of symptoms, PCOS is classified into four phenotypes according to the Rotterdam criteria: phenotype A (Oligo-anovulation, OA, clinical/biochemical hyperandrogenism HA, and the appearance of polycystic ovary PCO on ultrasonography); phenotype B (HA with OA only); phenotype C (HA with PCO only); and phenotype D (OA with PCO only). By including polycystic ovaries morphology (PCOM) as a criterion and permitting a diagnosis based on two of the three criteria, the Rotterdam criteria broadened the scope of the NIH criteria for PCOS diagnosis symptom.
As a metabolic abnormality, PCOS is widely recognised for its increased vulnerability to insulin resistance, poor glucose tolerance, obesity, type 2 diabetes, and dyslipidaemia4-6. Individuals with PCOS may experience serious long-term health consequences due to this metabolic abnormality, such as an elevated risk of cardiovascular diseases (CVDs) and other comorbidities. Roughly 70 per cent of females afflicted with PCOS possess at least one biochemical indicator denoting dyslipidaemia7. Dyslipidaemia, which is associated with abnormal levels of lipids, is common in PCOS and linked with an elevated risk of CVD.
Women with PCOS are usually found to have lipid abnormalities, like lower levels of high-density lipoprotein and raised levels of triglycerides (TGs) and low-density lipoprotein, which are quite concerning7. These lipid abnormalities raise the risk of diabetes, heart disease, and stroke by contributing to the development of metabolic syndrome.
Among the multifaceted metabolic disturbances associated with PCOS, alterations in lipid metabolism have gained considerable attention in recent research8. Thus, understanding the lipid profile in different phenotypes of PCOS is crucial for identifying specific metabolic disturbances and developing targeted management approaches. Therefore, this systematic review and meta-analysis aims to evaluate lipid changes, including total cholesterol, high-density lipoprotein-cholesterol (HDL-C), TGs, and low-density lipoprotein-cholesterol (LDL-C), in different phenotypes of PCOS compared with healthy controls in a representative sample of reproductive-aged women.
Materials & Methods
Search strategy
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards were followed for this study9. Worldwide published studies between January 2010 and December 2024 were systematically searched and assessed using electronic databases, including Science Direct, PubMed, Google Scholar, and Web of Science-Science Citation Index, where PCOS phenotypes were compared with controls. The search keywords used included ‘PCOS phenotype’, ‘Polycystic ovary syndrome phenotype’, ‘lipid levels’, and ‘lipid profile’. The study was restricted to human research and English language publications. Additionally, we manually searched the reference pages of relevant publications.
Selection criteria
The inclusion criteria for this review encompassed several key aspects. Firstly, only full-text articles were considered, while abstracts were excluded from the analysis. Secondly, observational studies, including cross-sectional, case-control, and cohort studies, were included in the review. Thirdly, the focus was on women of reproductive age, typically between the ages of 18 and 45, who were diagnosed with PCOS using either the Rotterdam criteria or the standard diagnostic criteria. Additionally, studies were required to report mean and standard deviation data or provide sufficient information to extract these parameters for both cases and controls.
The exclusion criteria for this review were also defined; pre-clinical research, review papers, case reports, and commentaries were removed from consideration. Furthermore, studies lacking a control group or comprising only a single study group without controls were not included in the review. Studies involving participants with conditions other than PCOS, such as androgen-secreting tumours, congenital adrenal hyperplasia, or those undergoing medications affecting androgen metabolism or lipid levels, were also excluded (Fig 1).
- Flowchart showing the number of papers retrieved by individual searches, and the number of included and excluded papers.
Data extraction and quality control assessment
Data such as the authors’ names, country of origin, publication year, study type, population characteristics, diagnostic criteria used, the total number of participants in both PCOS and control groups, and the outcomes observed in these groups were extracted from each study. The quality of studies included in this systematic review was assessed using the Newcastle-Ottawa Scale (NOS). This scale uses a ‘star system’ to assess the quality of studies across three main perspectives in three primary areas: the process of selecting study groups, comparability between the groups, and identifying the relevant outcome. Two independent reviewers (NM and PM) screened the titles and abstracts of all retrieved articles based on predefined inclusion and exclusion criteria. Full-text articles deemed potentially eligible were then independently assessed. In case of disagreement between the two reviewers, the opinion of a third reviewer (JK) was sought, and the quality assessment score was decided by consensus among the three reviewers. The NOS assessment instrument consists of nine questions to assess different study components. Each study was graded into different quality levels, based on the number of ‘star’ responses it received to these questions. Studies were classified as ‘poor’ if they received fewer than five ‘star’ ratings. Those with five to six ‘star’ ratings were considered ‘moderate’ quality, while studies with more than seven ‘star’ ratings were deemed ‘high quality’ (Table)10-22. Begg’s funnel plot was employed to assess the potential publication bias. This method helps detect asymmetry in the distribution of studies, which may suggest bias due to unpublished negative or non-significant results (Supplementary Fig. 1-4).
| Studies | Selection | Comparability | Outcomes |
|---|---|---|---|
| Yilmaz et al22, 2011 | *** | ** | *** |
| Cupisti et al14, 2011 | **** | ** | *** |
| Melo et al18, 2011 | ***** | ** | *** |
| Moro et al19, 2012 | **** | ** | *** |
| Ates et al10, 2013 | **** | * | ** |
| Tehrani et al20, 2014 | **** | ** | *** |
| Bagir et al11, 2015 | **** | * | ** |
| Jamil et al16, 2015 | **** | ** | *** |
| Praveen et al17, 2018 | ***** | ** | *** |
| Tripathy et al21, 2018 | ***** | ** | *** |
| Enrico Carmina et al12, 2019 | *** | ** | *** |
| Dadachanji et al15, 2020 | **** | ** | *** |
| Çintesun et al13, 2021 | **** | ** | *** |
Quality of selection for case/control (minimum 1 - maximum 4 stars); Comparability(minimum 0 - maximum 2 stars); Exposure(minimum 1 - maximum 3 stars).Quality of selection adapted for cross-sectional/cohort studies (minimum 0 - maximum 5 stars); Comparability (minimum 0 - maximum 2 stars); outcome (minimum 0 - maximum 3 stars). The stars above represent the score for a particular quality. One star represents one score
Data analysis
Information was gathered and organised into a table. Mean differences (MDs) and 95% confidence intervals (CIs) were calculated for each continuous outcome (Total Cholesterol, HDL-C, TGs, and LDL-C) for all included studies. The heterogeneity across all included studies was measured using the I2 statistic. The I2 statistic values were categorised as low (<25%), moderate (25-70%), and high (>70-100%), with a level of significance at P<0.05. Two types of effects modelling were used: fixed effects results were shown when the I2 value was less than 50 per cent, while random effects results were displayed when the I2 value was greater than 50 per cent. We constructed forest plots for LDL-C, TG, and HDL-C comparisons with the control using the Cochrane Review Manager (RevMan) version 5.4.
Results
Searches in different databases identified 1,771 articles. After eliminating duplicates and screening titles and abstracts, 182 full-text papers were evaluated for eligibility. Ultimately, 13 studies were included in the meta-analysis10-22. Although the inclusion criteria specified studies that used the Rotterdam diagnostic criteria, one study employed the Androgen Excess and PCOS Society (AE-PCOS) criteria14. This study was retained due to its high methodological quality and relevance to the review objectives, but this deviation is acknowledged as a limitation. The final dataset comprised 3,655 women with PCOS and 1,824 controls. PCOS participants were stratified by phenotype: 1,907 participants in phenotype A, 474 participants in phenotype B, 764 participants in phenotype C, and 510 participants in phenotype D.
The meta-analysis assessed four lipid parameters: total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), TGs, and low-density lipoprotein cholesterol (LDL-C). Since heterogeneity was observed among studies, a random-effects model was used to estimate the mean difference and its 95% CI. Demographic variables like Body mass index and Lipid profile of study groups according to different phenotypes have been presented in the supplementary table (Supplementary Table).
Women with the complete phenotype polycystic ovarian morphology plus hyperandrogenism plus ovulatory dysfunction (PCO+HA+O) showed higher levels of TC, LDL-C, and TG compared to women with other PCOS phenotypes. Figure 2 compares TC levels between PCOS phenotypes and control subjects, expressed in mg/dl. The values represent the mean difference in TC levels between each PCOS phenotype and the control group. Phenotype A PCOS individuals had a mean TC level that was 12.69 mg/dl more than that of controls (95% CI: 8.25, 17.13; Fig. 2A). Similarly, phenotype B individuals had a mean increase of 9.91 mg/dl (95% CI: 2.79, 17.03; Fig. 2B), and phenotype C individuals had a mean increase of 6.22 mg/dl (95% CI: 3.44, 9.01; Fig. 2C), relative to the control group. These differences in TC were statistically significant for PCOS individuals with phenotypes A, B, and C. In contrast, phenotype D individuals showed a non-significant increase of 3.26 mg/dl compared to control women (95% CI: 1.85, 8.37; P=0.21; Fig. 2D).
- Forest plot of comparison: Total cholesterol in PCOS phenotypes versus controls (mg/dl). (A) phenotype A vs. control; (B) Phenotype B vs. control; (C) Phenotype C vs. control; and (D) Phenotype D vs. control.
Figure 3 compares HDL-C (mg/dl) levels between women with different PCOS phenotypes and healthy controls. Compared to control participants, women with PCOS phenotypes A, B, and D exhibited significantly lower HDL-C levels by 4.94 mg/dl (95% CI: 2.66, 7.22; Fig. 3A), 5.64 mg/dl (95% CI: 2.13, 9.15; Fig. 3B), and 4 mg/dl (95% CI: 0.21, 8.22; Fig. 3D), respectively. Phenotype C demonstrated the smallest mean difference, with HDL-C levels 3.4 mg/dl lower than controls (95% CI: 0.88, 5.92; Fig. 3C).
- Forest plot of comparison: High density lipoprotein cholesterol in PCOS phenotypes versus controls (mg/dl). (A) phenotype A vs. control; (B) Phenotype B vs. control; (C) Phenotype C vs. control; and (D) Phenotype D vs. control.
Figure 4 compares TG levels (mg/dl) between women with PCOS and healthy controls. All four PCOS phenotypes showed a statistically significant increase in mean TG levels compared to controls. Specifically, women with phenotypes A, B, C, and D had higher TG levels by 22.65 mg/dl (95% CI: 16.48, 28.82; Fig. 4A), 13.49 mg/dl (95% CI: 5.51, 21.46; Fig. 4B), 10.75 mg/dl (95% CI: 1.41, 20.1; Fig. 4C), and 11.73 mg/dl (95% CI: 1.86, 21.6; Fig. 4D), respectively, when compared to healthy women. Figure 5 illustrates that LDL-C levels were elevated in women with PCOS across all phenotypes compared to control subjects. The mean difference in LDL-C concentration was 12.94 mg/dl for phenotype A (95% CI: 7.5, 18.39; Fig. 5A), 10.36 mg/dl for phenotype B (95% CI: 0.24, 20.48; Fig. 5B), 7.17 mg/dl for phenotype C (95% CI: 1.44, 12.9; Fig. 5C), and 6.47 mg/dl for phenotype D (95% CI: 1.64, 11.3; Fig. 5D), all indicating higher LDL-C levels in PCOS women compared to controls.
- Forest plot of comparison: Triglyceride in PCOS phenotypes versus controls (mg/dl). (A) phenotype A vs. control; (B) Phenotype B vs. control; (C) Phenotype C vs. control; and (D) Phenotype D vs. control.
- Forest plot of comparison: Low density lipoprotein cholesterol in PCOS phenotypes versus controls (mg/dl). (A) phenotype A vs. control; (B) Phenotype B vs. control; (C) Phenotype C vs. control; and (D) Phenotype D vs. control.
Discussion
This systematic review and meta-analysis synthesised data from worldwide observational studies to assess lipid profile differences between women with PCOS across various phenotypes and control participants. Our findings demonstrate considerable variability in lipid levels among individual studies, underscoring the influence of PCOS phenotype severity, classified into four phenotypes A, B, C, and D, on lipid metabolism.
In our pooled analysis, 52.18 per cent of PCOS cases were classified as phenotype A, aligning with a previous review that reported approximately 60 per cent of PCOS patients in this category23. Phenotype A, defined by the simultaneous presence of OA, clinical and/or biochemical HA, and PCO morphology, is widely regarded as the most severe and metabolically adverse presentation of PCOS. However, there was some variability in phenotype distribution across studies. For instance, Sachdeva et al24 reported a high prevalence of phenotype A (67.7%) in a group of 164 Indian women with PCOS, followed by phenotypes C (17.7%), B (11%), and D (3.6%). Similarly, Tavares et al25 found that phenotype A was the most common subtype (54.1%), followed by phenotype D (19.8%) in their observational study of 111 women. These differences in prevalence may reflect regional, ethnic, or methodological variations across studies, but they consistently underscore the dominance of phenotype A in clinical settings.
In our study, the PCOS group had lower HDL and higher mean total cholesterol, LDL, and TGs than the control group. As is typical of the metabolic syndrome seen in PCOS, the phenotypes A, B, and D had lower HDL and greater TGs than the control group. The lipid profile of phenotype A indicates a more severe cardiovascular risk-prone subpopulation. This result was comparable to a study by Teharani et al20, who found that phenotype A had a more negative lipid profile. In contrast, the other phenotypes in their study did not differ substantially from the control.
Our research also showed that PCOS was associated with specific lipid patterns, including low levels of HDL-C, high TGs, high TC, and high LDL-C, which are consistent with the findings of studies by Tsouma et al26 and Ghaffarzad et al27. Although LDL-C is considered the main goal for reducing CVD risk due to the high incidence of metabolic disorders in women with PCOS, most authors have focused on variations in TGs and HDL-C. Still, other lipid changes have received relatively less attention8,28-30. In contrast, our study focused on all lipid changes, including TG, HDL-C, LDL-C, and TC. In terms of lipid profile indicators (TG, TC, LDL, and HDL), Cupisti et al14 study evaluated the variations among the eight most prevalent characteristics of metabolic alterations.
Several studies conducted in the past years found higher values of LDL-C in PCOS women31-35. The results indicate that different PCOS phenotypes have varying lipid levels, with an unfavourable lipid profile being linked to the hyperandrogenic phenotype. HDL levels were lower in all PCOS phenotypes than in controls.
After comparing lipid profiles of the PCOS and control groups, we identified that the PCOS group had lower HDL and higher TC, LDL, and TG values. The results presented by Pehlivanov and Orbetzova were consistent with these TG and LDL levels36. Our investigation further identified decreased HDL values in all PCOS phenotypic categories compared to the control group, consistent with some previous research37-39. One Korean study, in contrast to ours, discovered that HDL levels were greater in PCO+OA phenotypic groups than in control groups. In that study, PCO individuals with HA had higher fasting insulin and TG levels than those without HA40.
This meta-analysis concludes by pointing out that women with PCOS, especially those with phenotype A, have notable lipid abnormalities. Although its exact contribution to the development of CVD is unknown, dyslipidaemia characterised by increased TGs, cholesterol, and LDL-C levels and decreased HDL-C in PCOS women, can potentially play a significant role in its development.
The findings emphasise the need for phenotype-specific risk assessment and management strategies to address metabolic and cardiovascular risk in PCOS subjects. Understanding the distinct lipid profiles across PCOS phenotypes may help clinicians to tailor more effective preventive and therapeutic interventions, leading to personalised medicine in the treatment of PCOS and its related comorbidities. However, more research is needed to explore the underlying mechanisms driving these lipid abnormalities and their impact on cardiovascular health in PCOS patients.
Financial support & sponsorship
None.
Conflicts of Interest
None.
Use of Artificial Intelligence (AI)-Assisted Technology for manuscript preparation
The authors confirm that there was no use of AI-assisted technology for assisting in the writing of the manuscript and no images were manipulated using AI.
References
- Understanding variation in prevalence estimates of polycystic ovary syndrome: A systematic review and meta-analysis. Hum Reprod Update. 2018;24:694-709.
- [Google Scholar]
- Prevalence, phenotype and cardiometabolic risk of polycystic ovary syndrome under different diagnostic criteria. Hum Reprod. 2012;27:3067-73.
- [Google Scholar]
- Polycystic ovary syndrome: Etiology, pathogenesis and diagnosis. Nat Rev Endocrinol. 2011;7:219-31.
- [Google Scholar]
- Polycystic ovary syndrome: A complex condition with psychological, reproductive and metabolic manifestations that impacts on health across the lifespan. BMC Med. 2010;8:41.
- [Google Scholar]
- Metabolic risk in PCOS: Phenotype and adiposity impact. Trends Endocrinol Metab. 2015;26:136-43.
- [Google Scholar]
- Association of insulin resistance with lipid profile, metabolic syndrome, and hormonal aberrations in overweight or obese women with polycystic ovary syndrome. J Health Popul Nutr. 2015;33:157-67.
- [Google Scholar]
- Lipoprotein lipid concentrations and cardiovascular risk in women with polycystic ovary syn drome. J Clin Endocrinol Metab. 1985;61:946-51.
- [Google Scholar]
- Differences in dyslipidemia between American and Italian women with polycystic ovary syndrome. J Endocrinol Invest. 2008;31:35-41.
- [Google Scholar]
- The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: Explanation and elaboration. BMJ. 2009;339:b2700.
- [Google Scholar]
- Different phenotypes of polycystic ovary syndrome in Turkish women: Clinical and endocrine characteristics. Gynecol Endocrinol. 2013;29:931-5.
- [Google Scholar]
- Body mass index below obesity threshold implies similar cardiovascular risk among various polycystic ovary syndrome Phenotypes. Med Princ Pract. 2016;25:61-6.
- [Google Scholar]
- Characterization of metabolic changes in the phenotypes of women with polycystic ovary syndrome in a large Mediterranean population from Sicily. Clin Endocrinol (Oxf). 2019;91:553-60.
- [Google Scholar]
- Serum sclerostin level and its relation to subclinical atherosclerosis in the polycystic ovary syndrome phenotypes: A prospective controlled study. Turk J Obstet Gynecol. 2021;18:167-74.
- [Google Scholar]
- The different phenotypes of polycystic ovary syndrome: No advantages for identifying women with aggravated insulin resistance or impaired lipids. Exp Clin Endocrinol Diabetes. 2011;119:502-8.
- [Google Scholar]
- Elucidating the impact of obesity on hormonal and metabolic perturbations in polycystic ovary syndrome phenotypes in Indian women. PLoS One. 2021;16:e0246862.
- [Google Scholar]
- A case-control observational study of insulin resistance and metabolic syndrome among the four phenotypes of polycystic ovary syndrome based on Rotterdam criteria. Reprod Health. 2015;12:1-9.
- [Google Scholar]
- Evaluation of metabolic, hormonal and clinical parameters in different phenotypes of polycystic ovary syndrome: An observational study from a tertiary care centre in Eastern India. J Diabetes Metab Disord Control . 2018;5:195-200.
- [Google Scholar]
- The frequency of metabolic syndrome is higher among PCOS Brazilian women with menstrual irregularity plus hyperandrogenism. Reprod Sci. 2011;18:1230-6.
- [Google Scholar]
- CD4+CD28null T lymphocyte frequency, a new marker of cardiovascular risk: Relationship with polycystic ovary syndrome phenotypes. Fertility and Sterility. 2012;98:1609-15.
- [Google Scholar]
- The prevalence of polycystic ovary syndrome, its phenotypes and cardio-metabolic features in a community sample of Iranian population: Tehran lipid and glucose study. Front Endocrinol (Lausanne). 2022;13:825528.
- [Google Scholar]
- Metabolic risk assessment of Indian women with polycystic ovarian syndrome in relation to four Rotterdam criteria based phenotypes. Eur J Obstet Gynecol Reprod Biol. 2018;224:60-5.
- [Google Scholar]
- Anthropometric, clinical and laboratory comparison of four phenotypes of polycystic ovary syndrome based on Rotterdam criteria. J Obstet Gynaecol Res. 2011;37:1020-6.
- [Google Scholar]
- Oligoanovulation with polycystic ovaries but not overt hyperandrogenism. J Clin Endocrinol Metab. 2006;91:3922-7.
- [Google Scholar]
- Comparison of the different PCOS Phenotypes based on clinical metabolic, and hormonal profile, and their response to clomiphene. Indian J Endocrinol Metab. 2019;23:326-31.
- [Google Scholar]
- The prevalence of metabolic syndrome in the different phenotypes of polycystic ovarian syndrome. Rev Bras Ginecol Obstet. 2019;41:37-43.
- [Google Scholar]
- Lipid lipoprotein profile alterations in Greek infertile women with polycystic ovaries: Influence of adipocytokines levels. In Vivo. 2014;28:935-9.
- [Google Scholar]
- Correlation of serum lipoprotein ratios with insulin resistance in infertile women with polycystic ovarian syndrome: A case control study. Int J Fertil Steril. 2016;10:29-35.
- [Google Scholar]
- Prevalence and characteristics of the metabolic syndrome in women with polycystic ovary syndrome. J Clin Endocrinol Metab. 2005;90:1929-35.
- [Google Scholar]
- Screening women with polycystic ovary syndrome for metabolic syndrome. Obstet Gynecol. 2005;106:131-7.
- [Google Scholar]
- Impaired glucose tolerance, type 2 diabetes and metabolic syndrome in polycystic ovary syndrome: A systematic review and meta-analysis. Hum Reprod Update. 2010;16:347-63.
- [Google Scholar]
- Phenotypic variation in hyperandrogenic women influences the findings of abnormal metabolic and cardiovascular risk parameters. J Clin Endocrinol Metab. 2005;90:2545-9.
- [Google Scholar]
- A more atherogenic serum lipoprotein profile is present in women with polycystic ovary syndrome: A case-control study. J Clin Endocrinol Metab. 2008;93:470-6.
- [Google Scholar]
- Relationship between the triglyceride/high-density lipoprotein-cholesterol ratio, insulin resistance index and cardiometabolic risk factors in women with polycystic ovary syndrome. Endocrinol Nutr. 2009;56:59-65.
- [Google Scholar]
- Milder forms of atherogenic dyslipidemia in ovulatory versus anovulatory polycystic ovary syndrome phenotype. Hum Reprod. 2009;24:2286-92.
- [Google Scholar]
- Clinical significance of inflammatory markers in polycystic ovary syndrome: Their relationship to insulin resistance and body mass index. Dis Markers. 2009;26:163-70.
- [Google Scholar]
- Characteristics of different phenotypes of polycystic ovary syndrome in a Bulgarian population. Gynecol Endocrinol. 2007;23:604-9.
- [Google Scholar]
- Risk of metabolic complications in the new PCOS phenotypes based on the Rotterdam criteria. Fertil Steril. 2007;88:1389-95.
- [Google Scholar]
- Is the PCOS diagnosis solved by ESHRE/ASRM 2003 consensus or could it include ultrasound examination of the ovarian stroma? Hum Reprod. 2006;21:3108-15.
- [Google Scholar]
- The frequency of metabolic syndrome in women with polycystic ovaries at reproductive ages and comparison of different metabolic syndrome diagnostic criteria. J Turk Ger Gynecol Assoc. 2007;8:402-7.
- [Google Scholar]
- Clinical and biochemical characteristics of polycystic ovary syndrome in Korean women. Hum Reprod. 2008;23:1924-31.
- [Google Scholar]