Women with polycystic ovary syndrome who present with hyperandrogenemia, hyperinsulinemia, and insulin resistance appear to be at high risk of cardiovascular disease. Elevated levels of endothelin-1, a marker of vasculopathy, have been reported in insulin-resistant subjects with endothelial dysfunction. Male gender also seems to be an aggravating factor for cardiovascular disease. In this study we investigated endothelin-1 levels in women with polycystic ovary syndrome, and we evaluated the effect of an insulin sensitizer, metformin, on endothelin-1 levels. Plasma endothelin-1 levels were measured in 23 obese (mean age, 24.3 +/- 4.6 yr; body mass index, 35 +/- 5.6 kg/m(2)) and 20 nonobese women with polycystic ovary syndrome (24.1 +/- 3.6 yr; body mass index, 21.8 +/- 2.5 kg/m(2)) as well as in 7 obese and 10 nonobese healthy, normal cycling, age-matched women. Additionally, endothelin-1 levels were evaluated in a subgroup of women with polycystic ovary syndrome (10 obese and 10 nonobese) 6 months postmetformin administration (1700 mg daily). Our results showed that obese and nonobese women with polycystic ovary syndrome had higher levels of endothelin-1 compared with the controls [obese, 2.52 +/- 1.87 vs. 0.44 +/- 0.23 pmol/liter (by analysis of covariance, P < 0.02); nonobese, 1.95 +/- 1.6 vs. 0.43 +/- 0.65 pmol/liter (P < 0.009)]. All of the participating women with polycystic ovary syndrome (n = 43) when compared with the total group of controls (n = 17) demonstrated hyperinsulinemia (polycystic ovary syndrome, 24.5 +/- 19.6; controls, 11.2 +/- 3.4 U/liter; P < 0.03), lower glucose utilization (M40) during the hyperinsulinemic euglycemic clamps (3.4 +/- 2.4 vs. 5.6 +/- 1.75 mg/kg.min; P < 0.045, by one-tailed test), and higher levels of endothelin-1 (polycystic ovary syndrome, 2.52 +/- 1.87; controls, 0.44 +/- 0.23 pmol/liter; P < 0.02, analysis of covariance covariate for body mass index). A positive correlation of endothelin-1 with free T levels was also shown (r = 0.4, P = 0.002) as well as a negative correlation of endothelin-1 with glucose utilization (r = -0.3; P = 0.033) in the total studied population. Finally, after metformin therapy, endothelin-1 levels were significantly reduced in obese (endothelin-1 before, 3.25 +/- 2.2; endothelin-1 after, 1.1 +/- 0.9 pmol/liter; P < 0.003) and nonobese (endothelin-1 before, 2.7 +/- 2; endothelin-1 after, 0.7 +/- 0.4 pmol/liter; P < 0.01) women with polycystic ovary syndrome, with no change in body mass index. Moreover, after metformin therapy, hyperandrogenemia and hyperinsulinemia were normalized, and glucose utilization improved [obese before: total T, 0.9 +/- 0.15 ng/ml; fasting insulin, 22.2 +/- 12.1 U/liter; glucose utilization, 2.15 +/- 0.5 mg/kg.min; obese after: total T, 0.5 +/- 0.2 ng/ml; fasting insulin, 11.6 +/- 6 U/liter; glucose utilization, 4.7 +/- 1.4 mg/kg.min 9P < 0.003, P < 0.006, and P < 0.002, respectively); nonobese before: total T, 1 +/- 0.5 ng/ml; fasting insulin, 15.5 +/- 7.6 U/liter; glucose utilization, 3.4 +/- 0.7 mg/kg.min; nonobese after: total T, 0.8 +/- 0.5 ng/ml; fasting insulin, 9 +/- 3.8 U/liter; glucose utilization, 6 +/- 1.7 mg/kg.min (P < 0.04, P < 0.02, and P < 0.0008, respectively)]. In conclusion, our data clearly demonstrate that women with polycystic ovary syndrome, obese and nonobese, have elevated endothelin-1 levels compared with the age-matched control group. In addition, 6 months of metformin therapy reduces endothelin-1 levels and improves their hormonal and metabolic profile.