Risk of Insulin Resistance and Metabolic Syndrome in Women with Hyperandrogenemia: A Comparison between PCOS Phenotypes and Beyond

Abstract

Polycystic ovary syndrome (PCOS) is the most common endocrine disorder in premenopausal women, with a wide spectrum of possible phenotypes, symptoms and sequelae according to the current clinical definition. However, there are women who do not fulfill at least two out of the three commonly used "Rotterdam criteria" and their risk of developing type 2 diabetes or obesity later in life is not defined. Therefore, we addressed this important gap by conducting a retrospective analysis based on 750 women with and without PCOS. We compared four different PCOS phenotypes according to the Rotterdam criteria with women who exhibit only one Rotterdam criterion and with healthy controls. Hormone and metabolic differences were assessed by analysis of variance (ANOVA) as well as logistic regression analysis. We found that hyperandrogenic women have per se a higher risk of developing insulin resistance compared to phenotypes without hyperandrogenism and healthy controls. In addition, hyperandrogenemia is associated with developing insulin resistance also in women with no other Rotterdam criterion. Our study encourages further diagnostic and therapeutic approaches for PCOS phenotypes in order to account for varying risks of developing metabolic diseases. Finally, women with hyperandrogenism as the only symptom should also be screened for insulin resistance to avoid later metabolic risks.

Keywords: PCOS; Rotterdam criteria; free testosterone; hyperandrogenism; insulin resistance.

Figure 1

Flowchart highlighting the selection process and criteria used for including participants in the study. PCOS: Polycystic ovary syndrome; CAH: Congenital Adrenal Hyperplasia.

Figure 2

Boxplots showing hormonal and metabolic differences between the polycystic ovary syndrome (PCOS) phenotypes A–D and the one Rotterdam criterion (1RC) and control groups. *: p < 0.05; **: p < 0.01; ***: p < 0.001; Red horizontal lines depict upper reference limits for the respective parameter according to the reference laboratory; Circles in the figure depict outliers, stars above the colums represent extreme values; (a) Differences in body-mass-index (BMI); (b) Differences in total testosterone (TT, measured via immunoassay); (c) Differences in free testosterone (fTesto); (d) Differences in androstenedione (ASD); further ASD outliers were: 22.05 ng/mL (group A) and 24.70 ng/mL (group B); (e) Differences in dehydroepiandrosterone-sulphate (DHEA-S) (f) Differences in sex-hormone binding globulin (SHBG); (g) Differences in modified Ferriman-Gallwey (mFG) score; (h) Differences in homeostasis model-assessment for insulin resistance (HOMA-IR); further HOMA-IR outliers were: 29.21 (group A); 25.96 (group A); 22.67 (group B); 17.76 (group B) and 17.30 (group 1RC); (i):Differences in Matsuda index; further Matsuda outliers were: 135.26 (group A) and 145.07 (group B); Phenotype definitions: A: Hyperandrogenism (HA), Oligomenorrhea (OM) and Polycystic Ovarian Morphology (PCOM) present; B: HA and OM present; C: HA and PCOM present; D: OM and PCOM present; 1RC: Only one criterion (HA, OM or PCOM) present; Co: Control group (no Rotterdam criteria present).

Figure 3

Pie chart showing the symptom distribution among 75 women with only one Rotterdam criterion (group 1RC); PCOM: Polycystic ovarian morphology; OM: Oligo-/Amenorrhea; HA: Hyperandrogenism; biochem.: biochemical; clin.: clinical.