Effects of androgen excess and body mass index on endothelial function in women with polycystic ovary syndrome

Polycystic ovary syndrome (PCOS) is associated with endothelial dysfunction; whether this is attributable to comorbid hyperandrogen-ism and/or obesity remains to be established. Therefore, we 1) compared endothelial function between lean and overweight/obese (OW/OB) women with and without androgen excess (AE)-PCOS and 2) examined androgens as potential modulators of endothelial function in these women. The flow-mediated dilation (FMD) test was applied in 14 women with AE-PCOS (lean: n = 7; OW/OB: n = 7) and 14 controls (CTRL; lean: n = 7, OW/OB: n = 7) at baseline (BSL) and following 7 days of ethinyl estradiol supplementation (EE; 30 lg/day) to assess the effect of a vasodilatory therapeutic on endothelial function; at each time point we assessed peak increases in diameter during reactive hyperemia (%FMD), shear rate, and low flow-mediated constriction (%LFMC). BSL %FMD was attenuated in lean AE-PCOS versus both lean CTRL (5.2 +/- 1.5 vs. 10.3 +/- 2.6%, P < 0.01) and OW/OB AE-PCOS (5.2 +/- 1.5 vs. 6.6 +/- 0.9%, P = 0.048). A negative correlation between BSL %FMD and free testosterone was observed in lean AE-PCOS only (R2 = 0.68, P = 0.02). EE increased %FMD in both OW/OB groups (CTRL: 7.6 +/- 0.6 vs. 10.4 +/- 2.5%, AE-PCOS: 6.6 +/- 0.9 vs. 9.6 +/- 1.7%, P < 0.01), had no impact on %FMD in lean AE-PCOS (5.17 +/- 1.5 vs. 5.17 +/- 1.1%, P = 0.99), and reduced %FMD in lean CTRL (10.3 +/- 2.6 vs. 7.6 +/- 1.2%, P = 0.03). Collectively, these data indicate that lean women with AE-PCOS exhibit more severe endothelial dysfunction than their OW/OB coun-terparts. Furthermore, endothelial dysfunction appears to be mediated by circulating androgens in lean but not in OW/OB AE-PCOS, suggesting a difference in the endothelial pathophysiology of AE-PCOS between these phenotypes. NEW & NOTEWORTHY We present evidence for marked endothelial dysfunction in lean women with androgen excess polycys-tic ovary syndrome (AE-PCOS) that is 1) associated with free testosterone levels, 2) impaired relative to overweight/obese women with AE-PCOS, and 3) unchanged following short-term ethinyl estradiol supplementation. These data indicate an important direct effect of androgens on the vascular system in women with AE-PCOS. Our data also suggest that the relationship between andro-gens and vascular health differs between phenotypes of AE-PCOS.