In humans, steroid hormones circulate in the blood mainly bound to specific steroid transport proteins, namely corticosteroid-binding globulin (CBG) for cortisol and sex hormone-binding globulin (SHBG) for testosterone and estradiol. The binding activities of these proteins are believed to modulate the biodisposal of steroids to target cells. It has been shown in vitro that insulin is a potent inhibitor of both CBG and SHBG secretion by a human hepatoblastoma cell (HepG2) line. To further investigate this potential effect of insulin in vivo, we prospectively studied three groups of lean subjects, obese subjects, and obese subjects with glucose intolerance, all of whom were otherwise healthy. The three groups were comparable in sex and age, and in the two obese groups, body mass index, waist to hip ratio, and blood pressure were similar. Plasma total CBG concentrations (38.2 +/- 5.4 vs. 31.7 +/- 4.05 mg/L; P = 0.016) and glycosylated CBG levels (37.3 +/- 5.2 vs. 31 +/- 3.9 mg/L; P = 0.018) were significantly increased in obese subjects with glucose intolerance. Plasma CBG correlated positively with fasting glucose levels (r = 0.49; P = 0.002), hemoglobin A1c levels (r = 0.35; P = 0.03), and area under the curve of glucose after an oral glucose tolerance test (r = 0.45; P = 0.005) and correlated negatively with the insulin response to i.v. glucose (AIRg; -0.38, P = 0.02) as well as to oral glucose (r = -0.40; P = 0.01) challenge tests. CBG levels did not covariate with insulin sensitivity. Multiple linear regression analysis showed that only AIRg contributed to the variability of the CBG concentration (P = 0.03), explaining 41% of its variance. Morning cortisol levels did not differ between the groups and did not correlate to any of the glucose or insulin metabolism parameters. Because carbohydrate chains influence the biological activity and half-life of glycoproteins, we analyzed the migration profile of CBG by Western blot and the interaction of CBG with lectin, Con A. The results indicated that the CBG mol wt and interaction with Con A did not differ between lean and obese patients. These data favor the hypothesis that the inhibitory effect of insulin on CBG liver secretion might be relevant in vivo and therefore contribute to decrease CBG levels in obese patients with enhanced insulin secretion. In both men and women, SHBG levels correlated negatively with fasting glucose (r = -0.55; P < 0.0001) and hemoglobin A1c (r = -0.38; P = 0.02) and positively with insulin sensitivity (S(I); r = 0.65; P = 0.003 and r = 0.63; P = 0.007 in men and women, respectively), but not with insulin secretion. The disposition index (S(I) x AIRg) was significantly decreased in the obese, glucose-intolerant subjects, suggesting that AIRg was inadequate for their degree of insulin resistance. The disposition index correlated positively with plasma SHBG levels (r = 0.52; P = 0.001) and negatively with plasma CBG levels (r = -0.54; P = 0.001). Our data suggest that CBG is a marker of insulin secretion in a similar way as SHBG is a marker of insulin sensitivity. As high plasma CBG levels have been associated with increased incidence of type 2 diabetes, this important issue merits further investigations.