Correlation of plasma insulin and insulin-like growth factor-I with indices of androgen transport and metabolism in women with polycystic ovary syndrome.

OBJECTIVE: Polycystic ovary syndrome (PCOS) is said to be associated with hyperinsulinaemia. Insulin stimulates androgen production by ovarian tissue in vitro and previous studies have identified a positive correlation of insulin with androstenedione. The aim of the present study was to discover whether insulin levels correlate with clinical presentation and with markers of androgen transport and metabolism in women with PCOS. DESIGN: Within-group analysis of clinical and biochemical characteristics of a consecutive series of women with PCOS, focusing on correlations of plasma insulin with clinical presentation and androgens. Insulin levels were also compared with a control group of normal women. PATIENTS: Forty-seven women who presented with hirsutism, cycle abnormalities or both, with ultrasound proven PCOS, were recruited. Mean age was 26.6 +/- 0.7 years (mean +/- SEM), BMI 27.3 +/- 1.2 kg/m2. MEASUREMENTS: Plasma insulin levels were measured at 30-minute intervals for 3 hours following a 75 g glucose load. Blood was also taken for measurement of testosterone (T), androstenedione (A), free testosterone (fT), sex hormone binding globulin (SHBG) and insulin-like growth factor-I (IGF-I). Androsterone glucoronide (AG), a marker of peripheral androgen metabolism, was also measured. RESULTS: Neither basal insulin nor the sum of insulin measurements during the glucose tolerance test (sumINS) in women with PCOS were significantly different from a control group with normal ovaries. Within the PCOS group, basal insulin was greater in women with irregular cycles or amenorrhoea than in those with regular ovulatory menses (8.0 +/- 1.1 vs 3.1 +/- 1.5 mU/l, P less than 0.01) despite similarly raised androgen levels. Both basal insulin and sumINS correlated with BMI in women with PCO (r = 0.37, P less than 0.05 and r = 0.64, P less than 0.01 respectively) but not in controls. There was no significant correlation between insulin or IGF-I levels and T, A or AG despite a positive correlation of AG (but no other androgen) with BMI. SHBG showed an inverse correlation and fT correlated positively with sumINS (r = -0.51, P less than 0.01; r = 0.39, P less than 0.05). Regression analysis of each of the androgens on the other variables demonstrated no significant relationship between insulin and androgens. CONCLUSIONS: These data suggest that, in vivo, the major effect of insulin on androgen secretion is mediated by changes in SHBG rather than by direct stimulation of ovarian androgen production. Higher insulin concentrations in anovulatory compared with ovulatory women with hyperandrogenaemia may indicate that insulin resistance in the ovary contributes to the mechanism of anovulation in PCOS.