Increased luteinizing hormone pulse frequency in obese oligomenorrheic girls with no evidence of hyperandrogenism.

OBJECTIVE: To determine whether obese, nonhirsute adolescents with oligomenorrhea exhibit similar increased LH pulse secretion patterns compared with obese girls with polycystic ovary syndrome (PCOS). DESIGN: Prospective, observational study. SETTING: Tertiary university hospital. PATIENT(S): Nine obese girls with oligomenorrhea, 15 with PCOS, and 10 controls. INTERVENTION(S): Twenty-four-hour IV blood sampling for LH (every 10 minutes); measurement of steroid hormones (every 12 hours); and injection of leuprolide acetate (10 microgm/kg SC). MAIN OUTCOME MEASURE(S): Twenty-four-hour, wake, and sleep LH mean serum concentration, pulse frequency, amplitude; steroid hormones, including free androgen index (FAI); and pre- and post-leuprolide acetate 17-hydroxyprogesterone measurements. RESULT(S): Twenty-four-hour LH pulse frequency in oligomenorrheic girls (18.6 +/- 1.2) (mean +/- SE) was comparable to that in girls with PCOS (20.9 +/- 0.7) and greater than in normal girls (13.4 +/- 0.8). The pulse number during both sleep and wake was identical in oligomenorrheic and PCOS girls and significantly greater than that of normal girls. Mean 24-hour LH level, serum androgen levels, and FAI in oligomenorrheic girls were equivalent to those of normal controls and lower than those of PCOS girls. CONCLUSION(S): These preliminary results indicate that obese girls with oligomenorrhea exhibit increased LH pulse frequency in the absence of clinical and/or biochemical evidence of hyperandrogenism.

Ovarian imaging by magnetic resonance in obese adolescent girls with polycystic ovary syndrome: a pilot study.

OBJECTIVE: To determine whether magnetic resonance (MR) imaging can serve as a useful investigational tool in the assessment of the polycystic ovary as compared with transabdominal ultrasound (US) for obese adolescents with polycystic ovary syndrome (PCOS). DESIGN: Prospective observational study. SETTING: Tertiary university hospital. PATIENT(S): Eleven obese adolescents with PCOS. INTERVENTION(S): Pelvic MR and US imaging and blood sampling. MAIN OUTCOME MEASURE(S): Total ovarian volume and follicle count; mean follicle count per longitudinal cross-section; stromal area; biochemical correlations with measured ovarian parameters. RESULT(S): With MR, the mean (+/-SE) total follicle count (21.9 +/- 1.3) was significantly greater than that observed with US (5.5 +/- 1.7) and significantly correlated with total ovarian volume. Two-dimensional cross-sectional analysis of the ovary by MR revealed a significantly greater mean follicle count (12.1 +/- 0.8) compared with the results obtained by US (3.0 +/- 0.5). The stromal area (173.3 +/- 25.1 mm2) was approximately 35% of the total ovarian surface area. No correlations were observed between biochemical indices and measured ovarian parameters. CONCLUSION(S): In contrast to US, MR provides vastly greater delineation of the structural components of the ovary in obese girls with PCOS and thus can serve as an excellent investigational technique to assess the morphological transformation of the adolescent ovary.

Progesterone inhibition of the hypothalamic gonadotropin-releasing hormone pulse generator: evidence for varied effects in hyperandrogenemic adolescent girls.

Compared with normal women, adults with polycystic ovarian syndrome (PCOS) require higher progesterone (P) concentrations to inhibit GnRH (LH) pulse frequency, which contributes to persistently rapid GnRH pulses and elevated LH levels in PCOS. To explore the origin of this abnormality, we assessed hypothalamic sensitivity to P feedback in nine normal controls and 11 hyperandrogenemic (HA) adolescents. Subjects first underwent frequent blood sampling for 11 h to assess baseline LH pulse frequency. Thereafter, oral estradiol and micronized P were given for 7 d to achieve mean estradiol and P levels of 143 +/- 16 pg/ml (524 +/- 60 pmol/liter) and 7.8 +/- 0.7 ng/ml (24.9 +/- 2.3 nmol/liter), respectively. LH pulse frequency was then reassessed. On d 7, the slope of the percent reduction of LH pulses per 11 h as a function of the d 7 P concentration was less in the HA group compared with controls (P = 0.02) despite similar P levels. LH pulse frequency was suppressed in all NC (mean, 7.0 to 3.4 pulses/11 h), but was unchanged in six of the HA girls (mean, 8.3 to 7.5 pulses/11 h). In contrast, in the other five HA adolescents, P induced similar slowing of LH pulses to that seen in NC (mean, 10.0 to 5.0 pulses/11 h). Baseline free testosterone levels were similar in both HA groups; the only observed difference between these HA groups is that the P-suppressible subjects were all of Hispanic descent. These data suggest that hyperandrogenemia during adolescence is variably associated with decreased sensitivity to P, which may have a partially genetic basis.

Luteinizing hormone secretion is not influenced by insulin infusion in women with polycystic ovary syndrome despite improved insulin sensitivity during pioglitazone treatment.

It has been reported in women with polycystic ovary syndrome (PCOS) that LH secretion is not altered by insulin infusion. To determine whether insulin resistance may have precluded an effect of insulin, pulsatile LH secretion and gonadotropin responses to GnRH were examined in PCOS women (n = 9) before and after pioglitazone treatment (45 mg/d) for 20 wk in the presence and absence of a hyperinsulinemic euglycemic clamp (80 mU/m2.min). Frequent blood samples were obtained for 12 h (every 10 min) as well as during sequential administration of GnRH at doses of 2, 10, and 20 microg over 12 h. A significant (P < 0.05) improvement in insulin sensitivity was seen in the subjects after treatment. Mean LH levels, LH pulse frequency and amplitude, as well as gonadotropin responses to GnRH were not influenced by pioglitazone, either with or without insulin infusion. We conclude that in PCOS women, inappropriate gonadotropin release does not appear to be a consequence of hyperinsulinemia.

Enhanced granulosa cell responsiveness to follicle-stimulating hormone during insulin infusion in women with polycystic ovary syndrome treated with pioglitazone.

Women with polycystic ovary syndrome (PCOS) are known to exhibit insulin resistance with compensatory hyperinsulinemia. To determine the role of hyperinsulinemia on follicle function in PCOS, we examined 24-h estradiol (E(2)) responses to recombinant human FSH (r-hFSH), 75 IU, before and during insulin infusion both before and after administration of pioglitazone (30 mg/d) in seven PCOS women. Each subject underwent two 10-h hyperinsulinemic-euglycemic clamps at rates of 30 (low dose) and 200 (high dose) mU/m(2).min, respectively. During both low- and high-dose insulin infusions, E(2) responses to r-hFSH were unaltered compared with that observed in the absence of insulin. Pioglitazone administration for 5 months improved insulin sensitivity as indicated by significantly (P < 0.05) increased glucose infusion rates during the clamp studies. At 3 months of treatment, r-hFSH-stimulated E(2) responses were not different from those observed before treatment. With pioglitazone treatment, E(2) responses to r-hFSH remained unchanged during low-dose insulin infusion, whereas a highly significant (P < 0.02) increased response was noted with the high-dose hyperinsulinemic-euglycemic clamp. In addition to a greater magnitude of response, peak levels of E(2) were sustained longer compared with that seen before treatment. The data indicate that granulosa cell responsiveness to FSH was enhanced by insulin after improved insulin sensitivity induced by pioglitazone. These findings are consistent with the possibility that PCOS granulosa cells are insulin resistant.

Increased luteinizing hormone secretion in women with polycystic ovary syndrome is unaltered by prolonged insulin infusion.

In PCOS women with insulin resistance, hyperinsulinemia may contribute to inappropriate gonadotropin secretion. To determine whether insulin influences gonadotropin release in PCOS, pulsatile LH secretion and gonadotropin responses to GnRH were evaluated before (phase 1) and during (phase 2) insulin infusion. In phase 1, 11 PCOS and 9 normal women on separate days underwent 1) frequent blood sampling (q 10 min) for 12 h and 2) gonadotropin stimulation by successive doses of GnRH, 2 microg, 10 microg, and 20 microg, administered i.v. at 4 h intervals over a continuous 12 h. In phase 2, studies were repeated 2 h after initiation of a 12-h hyperinsulinemic-euglycemic clamp (80 mU/m(2).min). Administration of insulin to both groups failed to alter mean serum gonadotropin concentrations, LH pulse frequency, or LH pulse amplitude. Moreover, gonadotropin responses to GnRH were unchanged by insulin infusion. In PCOS and normal women, a significant reduction of serum androstenedione was associated with insulin administration, whereas no differences were noted for the remaining androgens and estrogens measured. These findings demonstrated that in PCOS women, LH secretion and gonadotropin responses to GnRH were not influenced by insulin administration. Insulin infusion had little effect on steroid hormone production with the possible exception of androstenedione. These results suggest that inappropriate LH secretion in PCOS is not a direct consequence of insulin resistance and compensatory hyperinsulinemia.