Hyperandrogenemia and Altered Day-night LH Pulse Patterns

Summary

Participation Requirements

Description

This is a randomized, placebo-controlled, double-blinded crossover study to test the following hypothesis: In mid- to late pubertal girls with hyperandrogenism (HA), acute progesterone suppression of waking LH pulse frequency is greater after 2 weeks of spironolactone pretreatment compared to after ...

This is a randomized, placebo-controlled, double-blinded crossover study to test the following hypothesis: In mid- to late pubertal girls with hyperandrogenism (HA), acute progesterone suppression of waking LH pulse frequency is greater after 2 weeks of spironolactone pretreatment compared to after 2 weeks of placebo pretreatment. We will only study mid- to late pubertal girls with HA (i.e., girls who would be candidates for therapeutic spironolactone use). Subjects will complete two 18-hour Clinical Research Unit (CRU) admissions in separate menstrual cycles. Subjects will be randomized to be pretreated for 2 weeks with either oral spironolactone (50 mg twice daily) or placebo prior to the first CRU admission. Immediately before and during each CRU admission, oral micronized progesterone (0.8 mg/kg/dose) will be given at 0700, 1500, 2300, and 0700 h. During each CRU admission, blood will be obtained every 10 minutes through an indwelling iv catheter from 1800 to 1200 h. This will allow full characterization of pulsatile LH secretion in addition to other hormone measurements. Formal polysomnography will be performed during CRU admissions. A second CRU admission (performed at least 2 months later given blood withdrawal limits) will be identical to the first except that placebo pretreatment will exchanged for spironolactone pretreatment or vice versa (treatment crossover). The primary endpoint is LH pulse frequency while awake. (LH pulse frequency while asleep is an important secondary endpoint.) The wake LH pulse frequency data from the spironolactone and placebo admissions will be analyzed via a hierarchical linear mixed model (HLMM). The admission (spironolactone vs. placebo) will represent the fixed effect factor of the HLMM. Random effects will be utilized to account for the hierarchical variance-covariance structure of the two-period cross-over design. Wake LH pulse frequency in response to exogenous progesterone will be compared between the spironolactone admission and the placebo admission via a linear contrast of the HLMM least squares LH pulse frequency means. A similar analysis will be performed for sleep-related LH pulse frequency. Using published and preliminary data, we determined that, if 16 mid- to late pubertal girls with HA complete both admissions, we should have at least an 80% chance of detecting a 0.35 pulse/hour mean within-subject difference in wake LH pulse frequency between the spironolactone and placebo admissions with a two-sided false positive rejection rate of no more than 0.05.

Tracking Information