ASSOCIATION BETWEEN PHENOTYPES AND METABOLIC ABNORMALITIES IN WOMEN WITH PCOS

Association between phenotypes and metabolic abnormalities in women with polycystic ovary syndrome Background. Polycystic ovary syndrome (PCOS) is a heterogeneous disorder. According to the current Rotterdam criteria (Fertil Steril 2004;81:19) it may be diagnosed according to the presence of different combinations of three features, namely clinical and/or biochemical hyperandrogenism, chronic oligo-anovulation and ultrasonographic appearance of the ovaries. However, it is still under debate whether subjects diagnosed by these different combinations truly represent different phenotypes of the same condition. Many PCOS women show insulin resistance and other metabolic alterations, which are considered a relevant health issue in this condition. Aim. The aim of this study was to assess whether the different elements used in PCOS diagnosis similarly cluster with insulin resistance and metabolic abnormalities. Subjects and Methods. This study included 114 women with PCOS (mean age+SD 24.1+5.6 yr, BMI 30.1+8.6 kg/m2), diagnosed according to the Rotterdam criteria, and 35 normal-weight healthy controls. Hyperandrogenism was assessed by clinical examination and serum testosterone and SHBG assay, with calculation of free testosterone concentrations by the Vermeulen formula (J Clin Endocrinol Metab 1999;84:3666). In women with more than 8 menses per year, oligo-anovulation was assessed by luteal phase measurement of serum progesterone. Ovarian ultrasonography was carried out, when possible, with a transvaginal approach, recording the number of the follicles and their diameters, and the ovarian volume in both ovaries. In addition, insulin sensitivity was measured by the glucose clamp technique, uterine arteries pulsatility and resistance indices were measured by Doppler analysis, glucose tolerance was assessed by 75g OGTT. Serum lipids and uric acid were also measured. Results. Among these PCOS women, 80% had hyperandrogenism, 88% had oligo-anovulation and 90% had polycystic ovaries. Insulin induced glucose utilization in the clamp studies was reduced in 65% of subjects and the metabolic syndrome was diagnosed, according to the IDF 2009 criteria (Circulation 2009;20:1640), in 36% of them. A number of anthropometric, metabolic and endocrine features were different between PCOS women and controls. When BMI-matched groups were compared, many of these differences were no longer statistically significant. However, fasting glucose and insulin sensitivity, as well as the expected differences in serum androgens, remained statistically different between the groups. Insulin sensitivity was significantly lower in hyperandrogenic than in non-hyperandrogenic PCOS women. Similarly, other metabolic features were different between these two subgroups of patients. However, both oligo-anovulation and the ultrasonographic features of the ovaries were not associated with the metabolic features. Inclusion in the analysis of BMI, as a covariate, reduced but did not eliminate the association between hyperandrogenism and insulin resistance. At Doppler analysis, both the pulsatility index and the resistance index of the uterine arteries were significantly higher in PCOS women than in controls (both p<0.001) and these differences were maintained when BMI-matched subjects were compared. These features correlated with SHBG and free testosterone levels, whereas no relationships were found with the metabolic features. Conclusions. Insulin resistance and the metabolic syndrome are common findings in young PCOS women. Fat mass excess is associated with both hyperandrogenism and the metabolic abnormalities, but it does not entirely account for the relationship between androgen excess and insulin resistance. Among the elements currently used for diagnosing PCOS, only hyperandrogenism is associated with insulin resistance and the other metabolic abnormalities of these women, whereas oligo-anovulation and polycystic ovaries are not. These findings suggest that non-hyperandrogenic PCOS women may have a less severe clinical condition. PCOS is also characterized by Doppler flussimetric alterations of the uterine arteries, which are associated with hyperandrogenism, but not with the anthropometric and metabolic abnormalities.