Polycystic ovarian morphology with regular ovulatory cycles: insights into the pathophysiology of polycystic ovarian syndrome.

To determine the relevance of polycystic ovarian morphology (PCOM) to the pathophysiology of polycystic ovarian syndrome (PCOS), biochemical features associated with PCOS were examined in 68 women with an established history of regular ovulatory cycles and no clinical evidence of hyperandrogenism. Ovarian morphology was objectively assessed by pelvic ultrasound. LH, FSH, estradiol (E(2)), testosterone (T), androstenedione (Delta(4)A), SHBG, and dehydroepiandrosterone sulfate (DHEAS) were measured at baseline in the early follicular phase (EFP) in all subjects. LH, FSH, E(2), and progesterone (P(4)) were then measured daily for a complete menstrual cycle in 16 women with normal ovarian morphology and in 26 women with PCOM. T, Delta(4)A, SHBG, and DHEAS levels were measured in pools of three daily samples in each of the EFP, midcycle, and midluteal phases. An additional 26 normal women (13 with normal ovarian morphology and 13 with PCOM) were studied in the EFP to assess pulsatile LH secretion, insulin and glucose levels, and the ovarian response to human chorionic gonadotropin. At baseline, there were no differences in body mass index or hirsutism scores between women with PCOM and normal ovaries. In daily samples across the menstrual cycle LH, FSH, E(2), and P(4) did not differ between women with PCOM and those with normal ovaries, and there was no difference in LH pulse amplitude or frequency in the EFP frequent sampling studies. In women with PCOM, T (P < 0.01), free T (P < 0.005), and DHEAS (P < 0.01) levels were higher at baseline in the EFP, and SHBG was lower (P < 0.05). Differences in Delta(4)A did not reach significance (P = 0.14). T, free T, Delta(4)A, and DHEAS were also increased in PCOM across the menstrual cycle (P < 0.05). In addition, 17-hydroxyprogesterone (P < 0.02), Delta(4)A (P < 0.01), and T (P < 0.01) responses to human chorionic gonadotropin were greater in women with PCOM. Fasting glucose was not different between the two groups, but fasting insulin was higher (P < 0.02) in PCOM women as was insulin resistance calculated from homeostatic model assessment (P < 0.01). These studies demonstrate that PCOM in nonhirsute women with documented ovulatory cycles is associated with normal E(2), P(4), and gonadotropin dynamics, but higher androgen and insulin levels and lower SHBG levels. Taken together, these findings suggest that PCOM with ovulatory cycles exists as a discrete entity, represents the mildest form of ovarian hyperandrogenism, and is associated with greater insulin resistance than in women with normal ovarian morphology. The absence of any neuroendocrine abnormality in women with PCOM and ovulatory cycles suggests that gonadotropin dysfunction is not required for increased androgen secretion, but may be critical for development of the anovulatory disorder associated with PCOS.