Disorders of the Glucose Metabolism Correlate With the Phenotype and the Severity in Women With Polycystic Ovary Syndrome

Josef van Helden; Osman Evliyaoglu; Andreas Küberl; Ralf Weiskirchen


Clin Endocrinol. 2020;93(1):44-51. 

In This Article

Abstract and Introduction


Background: Different polycystic ovary syndrome (PCOS) phenotypes are correlated with different clinical severity levels. Insulin resistance correlates with higher severity. In a retrospective study, 130 patients with polycystic ovary syndrome were examined for insulin resistance. The aim of the study was to investigate relationships between glucose metabolism and different PCOS phenotypes and to identify biomarkers or combinations thereof to obtain information on the type of metabolic disorder or the severity of PCOS.

Methods: A total of 130 patients with PCOS were included in the study. Biometric data such as weight, height, cycle day and cycle length were compared with glucose metabolism parameters such as fasting glucose, insulin before and 60 and 120 minutes after 75 g glucose intake, intact proinsulin, C-peptide and ovarian function parameters including Anti-Müllerian hormone (AMH) and the soluble AMH receptor (sAMHR2). The parameters were correlated, and their diagnostic performance with respect to different expressions of PCOS was evaluated.

Results: The biomarkers of impaired glucose metabolism showed strong significant difference in HOMA Index, adiponectin, proinsulin and body mass index (BMI) and Insulin levels in 0-60-120 minutes of glucose tolerance test but also with parameters of ovarian function as AMH, AMH z -score sAMHR2, and sAMHR2/AMH ratio. A strong correlation between sAMHR2 and adiponectin (r = .818, P < .0001) was found indicating a relationship between the degree of glucose metabolic impairment and ovarian function.

Conclusions: The parameters glucose, insulin, insulin 60 minutes after intake of 75 g glucose and adiponectin or sAMHR2 enable a biochemical classification of PCOS patients that correlates with morphological PCOS phenotypes. By determining biomarkers, it is possible to classify PCOS patients into subgroups that correlate with different PCOS phenotypes and the clinical severity.


Insulin resistance is a common finding in women with polycystic ovary syndrome (PCOS). The reduced insulin efficacy is a crucial pathomechanism in the desquamation of type 2 diabetes. If insulin resistance is known, targeted therapy can be used to increase the probability of conception in PCOS patients. However, not all patients are equally affected depending on their phenotype.[1] The gold standard of diagnostics is without doubt the hyperinsulinemic euglycemic clamp.[2] Unfortunately, this is a complicated and time-consuming examination technique that also requires experienced operators and is therefore not suitable for routine examinations. This is why different surrogate indices of plasma glucose and insulin have established themselves in practice either on fasting or after oral glucose administration.[3] However, these indices only have a good agreement with the glucose clamp in overweight patients, whereas many patients with insulin resistance are not identified in normal weight patients, regardless of which method is used.[4] Insulin resistance is not a disease in itself, but it is involved in the development of many pathological changes.[5] Among other things, it is thought to play a pathogenetic role in PCOS.[6] The PCOS is divided into four different phenotypes (A-D).[7] The most common is phenotype A, in which the presence of polycystic ovaries is combined with chronic anovulation and hyperandrogenemia. Phenotype B consists of presence of PCOS and oligo/amenorrhoea, while phenotype C is the combination of PCOS with hyperandrogenemia and phenotype D is just the presence of PCOS morphology. All types of PCOS are associated with normal or elevated BMI and different percentages of insulin resistance.[8] This led to the conclusion that different phenotypes should also be screened differently in order to effectively treat the different metabolic dysfunctions,[9] which in turn should lead to an improvement in ovarian function. A biomarker panel consisting of intact proinsulin, C-reactive protein (CRP) and adiponectin was proposed to assess beta-cell function, insulin sensitivity and chronic systemic inflammatory response.[10] Adiponectin is an adipokine that, in addition to its antiatherogenic, antidiabetic, anti-inflammatory and insulin-sensitizing effects,[11] also has influences on ovarian function[12] and GnRH release in the hypothalamus.[13] It is presented in women with PCOS in decreased blood concentration.[14] In a previous study, adiponectin was not suitable for predicting the severity of a PCOS in order to assess the clinical severity.[15] The parameter with the highest predictive value for ovarian function so far is the Anti-Müllerian hormone (AMH). It has been shown that women with a PCOS value above the 90th percentile (corresponding to a z -score >1.90) have a 94.9% probability of showing an anovulatory PCOS.[16] However, the predictive value of AMH with regard to the onset of pregnancy is significantly lower. The best predictor here is still the patient's age of life. It therefore makes sense to search for further biomarkers or combinations of biomarkers that are able to make better statements in this respect. As a possible candidate, the soluble ectodomain of the AMH receptor 2 (sAMHR2) was measured for the first time. From the ratio of AMH and sAMHR2, it may be possible to draw further conclusions about the quality of folliculogenesis in patients with PCOS.[17,18]

To answer this question, we conducted a study on 130 patients with PCOS of different phenotypes and expressions, examining various parameters of glucose metabolism and correlating them with the parameters of ovarian function.