Progesterone-Mediated Inhibition of the GnRH Pulse Generator: Differential Sensitivity as a Function of Sleep Status

Progesterone-Mediated Inhibition of the GnRH Pulse Generator: Differential Sensitivity as a... Abstract Context During normal, early puberty, luteinizing hormone (LH) pulse frequency is low while awake but increases during sleep. Mechanisms underlying such changes are unclear, but a small study in early pubertal girls suggested that differential wake-sleep sensitivity to progesterone negative feedback plays a role. Objective To test the hypothesis that progesterone acutely reduces waking LH pulse frequency more than sleep-associated pulse frequency in late pubertal girls. Design Randomized, placebo-controlled, double-blinded crossover study. Setting Academic clinical research unit. Participants Eleven normal, postmenarcheal girls, ages 12 to 15 years. Intervention Subjects completed two 18-hour admissions in separate menstrual cycles (cycle days 6 to 11). Frequent blood sampling for LH assessment was performed at 1800 to 1200 hours; sleep was encouraged at 2300 to 0700 hours. Either oral micronized progesterone (0.8 mg/kg/dose) or placebo was given at 0700, 1500, 2300, and 0700 hours, before and during the first admission. A second admission, performed at least 2 months later, was identical to the first except that placebo was exchanged for progesterone or vice versa (treatment crossover). Main Outcome Measures LH pulse frequency during waking and sleeping hours. Results Progesterone reduced waking LH pulse frequency by 26% (P = 0.019), with no change observed during sleep (P = 0.314). The interaction between treatment condition (progesterone vs placebo) and sleep status (wake vs sleep) was highly significant (P = 0.007). Conclusions In late pubertal girls, progesterone acutely reduced waking LH pulse frequency more than sleep-associated pulse frequency. Differential wake-sleep sensitivity to progesterone negative feedback may direct sleep-wake LH pulse frequency changes across puberty. The hypothalamic-pituitary-gonadal axis is quiescent during childhood, and nocturnal amplification of luteinizing hormone [LH; mirroring gonadotropin-releasing hormone (GnRH)] pulse frequency and amplitude marks the beginning of puberty (1). Mechanisms underlying these marked sleep-wake differences in LH secretion remain unclear, but nocturnal amplification of LH release during puberty is clearly tied to sleep: it generally occurs within 1 hour of sleep and follows sleep reversal (2–4). As high and low GnRH pulse frequencies favor LH and follicle-stimulating hormone (FSH) production, respectively (1, 5, 6), these patterns may be critical for normal early pubertal development, and disruptions of such may be relevant to polycystic ovary syndrome (PCOS), a disorder characterized by abnormal gonadotropin secretion that frequently manifests around the time of puberty (7). Our group has demonstrated that in early pubertal girls, progesterone concentrations increase in the early morning, and we hypothesized that this may rapidly reduce morning GnRH pulse frequency, contributing to day-night differences in LH release (8). To test this notion, we assessed LH pulse secretion in early-to-midpubertal girls who did and did not receive exogenous progesterone in the evening, hypothesizing that progesterone would slow nocturnal LH pulse frequency. Whereas progesterone appeared to abolish late-evening LH pulse frequency (while awake), it did not alter nocturnal (sleep-related) LH pulse frequency (9). These data suggested that the pubertal GnRH pulse generator is differentially sensitive to progesterone inhibition based on sleep-wake status. However, this study involved a relatively small number of subjects receiving progesterone and limited surveillance of daytime LH secretion, and as some early pubertal girls exhibit little daytime LH secretion, it remained possible that some girls in the progesterone group would have exhibited no daytime LH secretion regardless. To test more rigorously the hypothesis that exogenous progesterone acutely reduces waking LH pulse frequency to a greater extent than sleep-associated LH pulse frequency in pubertal girls, we performed a randomized, placebo-controlled, double-blinded crossover study in late pubertal girls. Subjects and Methods Eleven healthy, late pubertal girls without hyperandrogenism were studied at the University of Virginia (UVA). Subjects were 12 to 15 years old, and all were postmenarcheal. Tanner breast stage was 3.9 ± 1.0 [mean ± standard deviation (SD)]. Eight subjects reported regular menstrual cycles. One subject, who was 1 year postmenarcheal, described menses every 2 months, and two subjects, who were postmenarcheal by <6 months, had not yet established a clear pattern. None had evidence for hormonal abnormalities (e.g., hyperandrogenism) or premature adrenarche. None had used medications known to affect the reproductive system, glucose metabolism, lipid metabolism, or blood pressure within 90 days of the study. Subject characteristics are shown in Table 1. Table 1. Subject Characteristics   n  Means  SD  Median  Range  Age, y  11  14.2  1.1  14.3  12.0–15.9  Bone age, y  10  15.0  1.3  15  13.5–17.0  Tanner breast stage  11  3.9  1.0  4  2–5  Years postmenarche  11  1.6  1.1  1  0–3  Cycle length, d  9a  33.6  10.8  30  24–60  BMI, kg/m2  11  27.5  11.5  23.6  18.6–57.4  BMI-for-age percentile  11  77.5  22.4  88  37–99  BMI z-score  11  1.09  1.05  1.15  −0.33 to 2.82  Waist circumference, cm  11  83.1  27.3  70.2  62.0–155.3  Hip circumference, cm  11  101.9  21.4  95.2  82.5–158.6  Waist-to-hip ratio  11  0.8  0.09  0.75  0.73–0.98  Body fat percentage  11  32.3  11.4  30.8  16.7–53.5  Total testosterone, ng/dLb  11  19.1  6.1  19  7.5–28.0  SHBG, nmol/Lb  11  42.4  16.3  46.7  18.7–76.7  Free testosterone, pg/mLb  11  3.3  1.5  3.0  1.5–5.4  Androstenedione, ng/dLb  11  176  96  157  68–376  DHEA-S, µg/dLb  11  104  59  83  50–247  Estradiol, pg/mLb  11  27.1  15.5  21.7  12.5–63.2  Fasting insulin, µIU/mLb  11  14.2  7.1  12.0  4.2–25.1  Fasting glucose, mg/dLb  10  94.3  7.2  94  86–111c  Hemoglobin A1c, %  10  5.2  0.3  5.2  4.8–5.6  IGF-1, ng/mLb  11  331  114  314  194–531    n  Means  SD  Median  Range  Age, y  11  14.2  1.1  14.3  12.0–15.9  Bone age, y  10  15.0  1.3  15  13.5–17.0  Tanner breast stage  11  3.9  1.0  4  2–5  Years postmenarche  11  1.6  1.1  1  0–3  Cycle length, d  9a  33.6  10.8  30  24–60  BMI, kg/m2  11  27.5  11.5  23.6  18.6–57.4  BMI-for-age percentile  11  77.5  22.4  88  37–99  BMI z-score  11  1.09  1.05  1.15  −0.33 to 2.82  Waist circumference, cm  11  83.1  27.3  70.2  62.0–155.3  Hip circumference, cm  11  101.9  21.4  95.2  82.5–158.6  Waist-to-hip ratio  11  0.8  0.09  0.75  0.73–0.98  Body fat percentage  11  32.3  11.4  30.8  16.7–53.5  Total testosterone, ng/dLb  11  19.1  6.1  19  7.5–28.0  SHBG, nmol/Lb  11  42.4  16.3  46.7  18.7–76.7  Free testosterone, pg/mLb  11  3.3  1.5  3.0  1.5–5.4  Androstenedione, ng/dLb  11  176  96  157  68–376  DHEA-S, µg/dLb  11  104  59  83  50–247  Estradiol, pg/mLb  11  27.1  15.5  21.7  12.5–63.2  Fasting insulin, µIU/mLb  11  14.2  7.1  12.0  4.2–25.1  Fasting glucose, mg/dLb  10  94.3  7.2  94  86–111c  Hemoglobin A1c, %  10  5.2  0.3  5.2  4.8–5.6  IGF-1, ng/mLb  11  331  114  314  194–531  To convert conventional to international system of units: total testosterone (ng/dL) × 3.467 (nmol/L); sex hormone-binding globulin (SHBG; µg/mL) × 8.896 (nmol/L); free testosterone (pg/mL) × 3.467 (pmol/L); androstenedione (ng/dL) × 0.131 (nmol/L); dehydroepiandrosterone sulfate (DHEA-S; µg/dL) × 27.211 (nmol/L); estradiol (pg/mL) × 3.671 (pmol/L); insulin (µIU/mL) × 7.175 (pmol/L); glucose (mg/dL) × 0.0555 (mmol/L); and insulinlike growth factor 1 (IGF-1; ng/mL) × 0.131 (nmol/L). Abbreviation: BMI, body mass index. a Summary data for cycle length does not include data for two girls who were <6 months postmenarcheal at the start of study. b Reported values were obtained from the placebo admission. Differences between placebo and progesterone admissions were assessed with Wilcoxon rank sum tests for androstenedione, DHEA-S, insulin, glucose, and IGF-1; all P values were >0.7. c Only one subject had a fasting glucose >100 during the placebo admission, but her fasting glucose levels from screening and the progesterone admission were 86 and 96, respectively. View Large Study procedures The Institutional Review Board at UVA approved all study procedures, which were in accordance with the ethical standards of the Helsinki Declaration of 1975, as revised in 2008. The study was registered with ClinicalTrials.gov (identifier NCT00929006). Informed assent and consent were obtained from all subjects and their custodial parents, respectively. Subjects underwent a detailed, physician-obtained medical history and physical examination, which included determination of pubertal stage (Tanner scale for breast development). As previously described (10), fasting blood tests were drawn at 0800 to 0900 hours to ensure good general health and to exclude hormonal abnormalities. All screening blood test results were unremarkable for each subject. Measures of adiposity included waist circumference and hip circumference, and BOD POD® was used to measure percent body fat. Plain x-ray of the left hand and wrist was obtained to assess bone age. Subjects underwent two separate admissions for frequent blood sampling (LH pulse analysis) in the UVA Clinical Research Unit (CRU; Fig. 1). All admissions occurred in the mid-to-late follicular phase (cycle days 6 to 11). Red blood cell counts, β-human chorionic gonadotropin, and plasma progesterone levels were checked 1 to 3 days before each overnight admission to exclude anemia, pregnancy, and the luteal phase, respectively. Figure 1. View largeDownload slide Study schematic. Vertical arrows indicate the times when either oral micronized progesterone (P4) or placebo (PBO) was given. PBO was exchanged for P4 or vice versa for the second admission (treatment crossover). Horizontal arrows indicate wake (1900 to 2300 and 0700 to 1100 hours) and sleep (2300 to 0300 and 0300 to 0700 hours) blocks. Figure 1. View largeDownload slide Study schematic. Vertical arrows indicate the times when either oral micronized progesterone (P4) or placebo (PBO) was given. PBO was exchanged for P4 or vice versa for the second admission (treatment crossover). Horizontal arrows indicate wake (1900 to 2300 and 0700 to 1100 hours) and sleep (2300 to 0300 and 0300 to 0700 hours) blocks. Subjects were randomized to receive either progesterone or placebo doses before (i.e., at 0700 and 1500 hours) and during (i.e., at 2300 and 0700 hours) the first overnight admission. Progesterone was administered as a cherry-flavored micronized progesterone suspension compounded by the UVA Investigational Pharmacy (US Food and Drug Administration investigational new drug 64,126); the placebo solution was identical, except that micronized progesterone was omitted. Each progesterone dose was 0.8 mg/kg body weight, which generally produces a mean progesterone level of 3 to 8 ng/mL when given three times daily. Subjects, CRU staff, and investigators were blinded to treatment condition. Subjects were admitted to the CRU at 1600 hours. Beginning at 1800 hours, blood was obtained through an indwelling, intravenous forearm catheter over an 18-hour period (1800 to 1200 hours): LH every 10 minutes; progesterone every 30 minutes; FSH, estradiol, and total testosterone every hour. Additional 1-mL samples taken at 1800, 0000, 0600, and 1200 hours were pooled for measurement of total testosterone by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Lights were extinguished at 2300 hours. Sleep was encouraged from 2300 to 0700 hours, but sleep was otherwise disallowed. Sleep periods were formally evaluated using wrist actigraphy (Motionlogger Basic-L; Ambulatory Monitoring, Ardsley, NY). Subjects fasted from 2300 to 0700 hours; at 0700 hours, an additional 4 mL was drawn for measurement of insulin, glucose, sex hormone-binding globulin (SHBG), dehydroepiandrosterone-sulfate, insulinlike growth factor 1, and androstenedione. Subjects were discharged after the final blood draw at 1200 hours. At least 2 months after the first admission, subjects returned for a second overnight admission identical to the first, except that exogenous progesterone was exchanged for placebo, or vice versa, in accordance with the crossover design. We enforced a 2-month minimum between admissions to allow adequate restoration of the subject’s red blood cell population. Hormonal measurements UVA’s Center for Research in Reproduction Ligand Assay and Analysis Core performed all hormone assays except for total testosterone by LC-MS/MS. Hormone concentrations measured below assay sensitivity were assigned the value of the assay’s sensitivity. LH was measured by chemiluminescence [Diagnostic Products, Los Angeles, CA; sensitivity, 0.1 IU/L; intra-assay coefficient of variation (CV) ≤3.3%; interassay CV ≤6.6%]. Each subject’s LH samples were analyzed in duplicate in the same assay. Other assay characteristics are described in detail in the Supplemental Materials. Before January 2015, sex steroids were measured by radioimmunoassay (RIA), as previously described (10), but in 2014, most steroid RIA kits used by the Center for Research in Reproduction were discontinued by the manufacturer and replaced by other immunoassays. Evaluations of replacement methods were conducted according to recommendations of the Endocrine Society Sex Steroid Assays Reporting Task Force (11). Each replacement method was carefully correlated to Diagnostic Products RIA, and RIA-equivalent values are presented herein. Total testosterone by LC-MS/MS was performed by the Mayo Clinic Laboratory [sensitivity 7 ng/dL; intra-assay CV ≤6.0%; interassay CVs 7.9% to 15.8% at 12 to 48.6 ng/dL (12)]. Total testosterone measurements by immunoassay and LC-MS/MS were highly concordant (Supplemental Materials). However, as the same LC-MS/MS testosterone assay was performed for all subjects in this study—in contrast to the use of different testosterone immunoassay methodologies over time—we elected to use LC-MS/MS results for formal analyses. Free testosterone was calculated from total testosterone and SHBG (13). Data analysis Data analyses related to pulsatile LH secretion was performed by a single investigator (C.R.M.) while blinded to treatment condition. As previously described (14), a computerized data reduction protocol (StdCurve) was used to provide a variance model for experimental measurement error for each LH concentration time series. LH pulses were then identified using the computerized pulse detection algorithm AutoDecon, a fully automated (i.e., nonsubjective), multiparameter deconvolution program (15). To limit false positives, only the following AutoDecon-identified pulses were included in subsequent analyses: (1) pulses with at least two peak values that were at least 10% higher than the preceding nadir or (2) pulses with at least one peak value that was at least 20% higher than the preceding nadir. Data derived from AutoDecon included the temporal locations of LH pulses, which were used to calculate LH pulse frequency, and LH pulse mass, an estimate of the amount of LH secreted by the pituitary during each pulse (a correlate of LH pulse amplitude). With the use of previously described methods (16), average LH interpulse intervals were calculated during four time blocks: 1900 to 2300, 2300 to 0300, 0300 to 0700, and 0700 to 1100 hours. For each time block, LH pulse frequency (expressed as pulses per hour) was then calculated by dividing 60 by the average interpulse interval. Our method of LH pulse-frequency determination for discrete time blocks is described in detail in Supplemental Materials. Wake LH pulse frequency was defined as the average of 1900 to 2300- and 0700 to 1100-hour blocks, and sleep LH pulse frequency was defined as the average of 2300 to 0300- and 0300 to 0700-hour blocks. LH pulse frequency was similarly calculated for actigraphy-determined sleep periods (between the first and last sleep epochs) and wake periods [weighted average of (1) frequency before first sleep and (2) frequency after last sleep]. Statistical analyses As the a priori primary statistical analysis, LH pulse frequency during wake and sleep periods of the placebo and progesterone admissions was analyzed in accordance with a two-period crossover design paradigm by way of linear-mixed (LM) models. The LM model was specified so that the variability in LH pulse frequency related to sequence order (progesterone vs placebo given first) and crossover period (first CRU admission vs second CRU admission) could be extracted from the variability in LH pulse frequency related to sleep status (sleep vs wake), treatment (progesterone vs placebo), and the sleep status by treatment interaction. With regard to hypothesis testing, the LM model covariance matrix was specified in the unstructured form, and linear contrasts of the least-square means were used to test both (1) if progesterone-induced changes in LH pulse frequency under the sleep and wake conditions were equal to zero and (2) if progesterone-induced changes in LH pulse frequency were the same under sleep and wake conditions. As preplanned secondary statistical analyses, LH pulse mass, mean LH, and mean FSH of the wake and sleep periods of the placebo and progesterone admissions were analyzed using LM models, as described previously. Changes in mean progesterone, estradiol, total testosterone, and free testosterone concentrations between admissions were analyzed via paired Student’s t tests. As post hoc analyses, we performed Wilcoxon signed-rank tests to assess for progesterone-related LH pulse-frequency differences (progesterone vs placebo admission) for each 4-hour time block (i.e., 1900 to 2300, 2300 to 0300, 0300 to 0700, and 0700 to 1100 hours). Spearman’s rank correlation coefficient (rs) was used to assess whether progesterone-induced changes in waking LH pulse frequency were associated with baseline free testosterone concentration. SAS version 9.4 (SAS Institute, Cary, NC) was used for all statistical analyses. A two-sided P ≤ 0.05 decision rule was used as the null hypothesis rejection criterion for all statistical tests. Data are presented as means ± standard error of the mean unless indicated otherwise. Subject-level data used for analyses are provided in Supplemental Materials. Results Sex steroid concentrations and actigraphy data Mean progesterone concentrations were higher during progesterone admissions compared with placebo admissions (6.5 ± 1.3 vs 0.4 ± 0.03 ng/mL; P = 0.001; Fig. 2). Mean estradiol was similar between admissions (27.7 ± 6.0 and 27.1 ± 4.7 pg/mL, respectively). Total testosterone by LC-MS/MS was lower during progesterone admissions compared with placebo admissions (15.3 ± 2.1 vs 19.1 ± 1.8 ng/dL; P = 0.047). SHBG concentrations were similar between progesterone and placebo admissions (40.8 ± 6.8 vs 42.4 ± 4.9 nmol/L, respectively; P = 0.757), as were calculated free testosterone concentrations using LC-MS/MS (2.7 ± 0.5 vs 3.3 ± 0.4 pg/mL, respectively; P = 0.219). Figure 2. View largeDownload slide Sex steroid changes between progesterone (P4) and placebo (PBO) conditions. Data points indicate the change in (a) mean P4, (b) estradiol, (c) total testosterone, and (d) free testosterone concentrations from the PBO to the P4 condition. Change in hormone concentration in individual subjects (○); these data are also summarized using box-and-whisker plots, which show median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). P values relate to the mean changes between the treatment conditions. E2, estradiol; T, testosterone. Figure 2. View largeDownload slide Sex steroid changes between progesterone (P4) and placebo (PBO) conditions. Data points indicate the change in (a) mean P4, (b) estradiol, (c) total testosterone, and (d) free testosterone concentrations from the PBO to the P4 condition. Change in hormone concentration in individual subjects (○); these data are also summarized using box-and-whisker plots, which show median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). P values relate to the mean changes between the treatment conditions. E2, estradiol; T, testosterone. Per wrist actigraphy, sleep duration (time from first sleep epoch to last sleep epoch) was 7.7 ± 0.7 and 7.7 ± 0.5 hours (means ± SD) during progesterone and placebo admissions, respectively; sleep latency (i.e., time from lights out to first sleep) was 19 ± 43 and 19 ± 33 minutes, respectively; and sleep efficiency (i.e., percentage of sleep duration occupied by sleep epochs) was 89 ± 8 and 92 ± 3%, respectively. Effect of progesterone on wake vs sleep-related LH pulse frequency Average waking LH pulse frequency was 26% lower during progesterone admissions compared with placebo admissions (0.74 ± 0.03 vs 1.00 ± 0.02 pulses/h; P = 0.019; Fig. 3). Average sleep-related LH pulse frequencies were 0.55 ± 0.02 and 0.64 ± 0.02 pulses per hour during progesterone and placebo admissions, respectively (P = 0.314). The interaction between treatment condition and sleep status was highly significant (P = 0.007). Whereas sleep status was an important predictor of LH pulse frequency (P < 0.001), sequence order and crossover period were not (P > 0.6 for both). Figure 3. View largeDownload slide LH pulse frequency as a function of sleep status and treatment condition. (a) Individual average LH pulse frequency under the placebo (PBO) and progesterone (P4) conditions during both (left) wake periods and (right) sleep periods. An individual subject’s data are represented by connected ○. Note that wake LH pulse frequency was lower with P4 than with PBO in most (10 of 11) subjects, whereas sleep-related LH pulse frequency was lower with P4 in less than one-half (five of 11) of subjects. (b) Mean LH pulse frequency (denoted by ●) in each treatment condition during (left) wake and (right) sleep blocks, with vertical lines denoting 95% confidence intervals. P values in each column relate to the change in LH pulse frequency between treatment conditions. P values relating to the interaction between treatment (PBO vs P4) and sleep status (wake vs sleep) are also shown. Figure 3. View largeDownload slide LH pulse frequency as a function of sleep status and treatment condition. (a) Individual average LH pulse frequency under the placebo (PBO) and progesterone (P4) conditions during both (left) wake periods and (right) sleep periods. An individual subject’s data are represented by connected ○. Note that wake LH pulse frequency was lower with P4 than with PBO in most (10 of 11) subjects, whereas sleep-related LH pulse frequency was lower with P4 in less than one-half (five of 11) of subjects. (b) Mean LH pulse frequency (denoted by ●) in each treatment condition during (left) wake and (right) sleep blocks, with vertical lines denoting 95% confidence intervals. P values in each column relate to the change in LH pulse frequency between treatment conditions. P values relating to the interaction between treatment (PBO vs P4) and sleep status (wake vs sleep) are also shown. Notably, LH pulse frequency for actigraphy-directed sleep periods correlated very highly with average LH pulse frequency from 2300 to 0700 hours (rs = 0.977, P < 0.0001), and results were essentially identical when actigraphy-estimated sleep and wake periods were used for analysis: LH pulse frequency during wake periods was lower with progesterone compared with placebo (P = 0.021), with no difference during sleep periods (P = 0.417), and the interaction between treatment condition and sleep status was highly significant (P = 0.007). Effect of progesterone on LH pulse mass, mean LH, and mean FSH LH pulse mass was higher with progesterone compared with placebo during both wake periods (5.32 ± 0.23 vs 2.44 ± 0.13 IU/L; P = 0.006) and sleep periods [7.47 ± 0.30 vs 3.75 ± 0.16 IU; P = 0.016; Fig. 4(a) and 4(b)]. Important predictors of LH pulse mass included sleep status (P = 0.001) but not sequence order or crossover period, and there was no interaction between treatment condition and sleep status. Figure 4. View largeDownload slide Secondary outcomes: LH pulse mass, mean LH, and mean FSH as a function of sleep status and treatment condition. Rows illustrate data for (a and b) LH pulse mass, (c and d) mean LH, and (e and f) mean FSH. (a, c, and e) (left) Individual mean data (denoted by connected ○) with placebo (PBO) and progesterone (P4) during wake blocks; (right) individual mean data during sleep blocks. (b, d, and f) (left) Mean data (denoted by ●) for PBO and P4 conditions during wake blocks; (right) mean data during sleep blocks. Vertical lines identify 95% confidence intervals for the mean. P values relate to changes between PBO and P4 conditions. Although not significant, P values relating to the interaction between treatment condition and sleep status are shown. Figure 4. View largeDownload slide Secondary outcomes: LH pulse mass, mean LH, and mean FSH as a function of sleep status and treatment condition. Rows illustrate data for (a and b) LH pulse mass, (c and d) mean LH, and (e and f) mean FSH. (a, c, and e) (left) Individual mean data (denoted by connected ○) with placebo (PBO) and progesterone (P4) during wake blocks; (right) individual mean data during sleep blocks. (b, d, and f) (left) Mean data (denoted by ●) for PBO and P4 conditions during wake blocks; (right) mean data during sleep blocks. Vertical lines identify 95% confidence intervals for the mean. P values relate to changes between PBO and P4 conditions. Although not significant, P values relating to the interaction between treatment condition and sleep status are shown. Although mean LH was approximately twofold higher with progesterone during waking and sleep hours, these differences were not statistically significant [Fig. 4(c) and 4(d)]. Mean FSH was similar under progesterone and placebo conditions during wake periods and sleep periods [Fig. 4(e) and 4(f)]. Mean LH and mean FSH were not predicted by sleep status, sequence order, or crossover period, and there was no interaction between treatment condition and sleep status for either. Post hoc analyses Compared with the placebo condition, progesterone was associated with significantly lower LH pulse frequency during both evening wake (1900 to 2300 hours) and morning wake (0700 to 1100 hours) time blocks, with no differences observed during either sleep-related time block [(Fig. 5(a)]. The Spearman’s rank correlation procedure did not disclose a significant correlation between (1) percent change in waking LH pulse frequency related to progesterone administration and (2) baseline free testosterone concentration in this cohort of late pubertal girls [rs = 0.291, P = 0.386; Fig. 5(b)]. Also of note, an appraisal of the data partitioned by obesity status suggested that study results were similar in obese and nonobese subjects (see Supplemental Materials). Figure 5. View largeDownload slide Post hoc analyses. (a) Mean LH pulse frequency with placebo (PBO) and progesterone (P4) in each 4-hour time block. Wake time blocks are 1900 to 2300 and 0700 to 1100 hours; sleep time blocks are 2300 to 0300 and 0300 to 0700 hours. Box-and-whisker plots denote the median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). (b) The relationship between the percent change in waking LH pulse frequency related to P4 administration and free testosterone measured during the PBO admission (calculated using SHBG and total testosterone measured by LC-MS/MS). Figure 5. View largeDownload slide Post hoc analyses. (a) Mean LH pulse frequency with placebo (PBO) and progesterone (P4) in each 4-hour time block. Wake time blocks are 1900 to 2300 and 0700 to 1100 hours; sleep time blocks are 2300 to 0300 and 0300 to 0700 hours. Box-and-whisker plots denote the median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). (b) The relationship between the percent change in waking LH pulse frequency related to P4 administration and free testosterone measured during the PBO admission (calculated using SHBG and total testosterone measured by LC-MS/MS). Discussion The current study provides compelling evidence that in late pubertal girls without hyperandrogenism, progesterone acutely suppresses waking LH pulse frequency to a greater extent than sleep-associated LH pulse frequency. These data support our earlier findings in early-to-midpubertal girls (9), and taken together, these two studies strongly suggest that the GnRH pulse generator in pubertal girls is differentially sensitive to progesterone inhibition based on sleep-wake status. We hypothesize that this phenomenon contributes to the evolution of sleep-wake LH pulse patterns across puberty in girls. As previously described, waking LH pulse frequency gradually increases across puberty, whereas sleep-associated LH frequency remains relatively constant: in early puberty, LH pulse frequency increases to ∼0.5 pulses/h during sleep, but by late puberty, LH pulse frequency decreases to ∼0.5 pulses/h during sleep (3, 10). Mechanisms underlying the evolution of such patterns are unknown. However, progesterone is a primary modulator of day-to-day GnRH pulse frequency slowing in women (17, 18), and contrary to girls with spontaneous puberty, age-matched girls with gonadal dysgenesis demonstrate no sleep-wake differences in LH pulse frequency (19), implying a role for sex steroids. Our previous study in early pubertal girls (9) suggested that exogenous progesterone quickly suppresses LH pulse frequency while awake but not during sleep. We thus hypothesized that waking GnRH frequency is regulated by a complement of neuronal inputs that are sensitive to progesterone negative feedback but that sleep-related GnRH pulse frequency is regulated by an ensemble of neuronal inputs that, as a unit, is relatively stereotyped and not readily influenced by low levels of progesterone (1). The results of our current study strongly support this hypothesis. We note that other studies have suggested differential control of sleep vs wake LH pulse frequency: in normal adult women studied in the late follicular phase, dietary calorie restriction preferentially reduced daytime LH pulse frequency (20, 21). According to our overall working model (1)—represented in schematic form in Supplemental Materials—waking and sleep-associated GnRH pulse frequency are regulated by different complements of neuronal inputs: the former is responsive to progesterone negative feedback, but the latter is relatively resistant to such inhibition. Very low waking GnRH pulse frequency in early puberty relates to exquisite sensitivity to progesterone negative feedback. [We suggest that in normal early pubertal girls (i.e., before the establishment of cyclic ovulation), high nocturnal frequency may be necessary to enhance LH secretion, which promotes ovarian sex steroid production, whereas a low daytime pulse frequency may be critical to maintain FSH secretion, which supports follicular development.] Androgen concentrations increase across puberty (10, 22, 23), and androgens antagonize progesterone negative feedback (24, 25): our working model holds that the gradual rise of testosterone across puberty reduces the ability of progesterone to inhibit waking GnRH pulse frequency, thus allowing a gradual increase in waking pulse frequency. However, outside of the luteal phase, sleep-associated GnRH pulse frequency remains relatively constant across puberty, as such inputs are not readily altered by low progesterone levels. We have also previously proposed that our working model may help explain the genesis of neuroendocrine abnormalities in girls with hyperandrogenemia and nascent PCOS (1): if pubertal initiation (GnRH pulse generator reawakening) occurs in the setting of relative hyperandrogenemia, then the higher androgen concentrations would antagonize progesterone inhibition of waking GnRH pulse frequency, leading to a rapid transition to high GnRH pulse frequency during both day and night. This pattern would exaggerate LH release and further promote androgen production while also limiting FSH synthesis and, thus, follicular development. We are currently pursuing studies to test these hypotheses. Whereas the post hoc analysis in the current study did not suggest a significant correlation between progesterone-induced changes in waking LH pulse frequency and free testosterone concentrations, we suggest that this may reflect relatively low numbers of subjects who were all within a narrow developmental range: inclusion of a greater number of subjects exhibiting a broader range of testosterone values may be required to demonstrate this putative relationship. Of interest, total testosterone by LC-MS/MS was 20% lower with progesterone administration compared with the placebo condition. Although progesterone reduced LH pulse frequency while awake, LH pulse mass simultaneously increased, and mean LH concentrations did not change. This may suggest that waking LH pulse frequency per se is an important determinant of ovarian androgen production, and it is consistent with the notion that an inability to suppress waking LH pulse frequency appropriately in early puberty may promote hyperandrogenemia and support a progression to PCOS. Our prior study in early pubertal girls suggested that progesterone abolishes waking LH pulse secretion within 3 to 7 hours (9), and our current study suggests that progesterone reduces waking LH pulse frequency by 26% within 12 to 16 hours in later puberty. In contrast, our previous studies in adult women suggest that progesterone suppresses waking LH pulse frequency more slowly (14, 26). We hypothesize that these apparently discordant observations reflect differences in androgen exposure: the effects of progesterone negative feedback are rapid and profound when androgen concentrations are very low, but they are less profound—or perhaps require longer periods of time—when androgen concentrations are higher. We also note that progesterone will suppress sleep-related LH pulse frequency when progesterone exposure is sufficiently high and long, as occurs in the luteal phase (27, 28). Progesterone administration given for 7 to 10 days, with or without concomitant estradiol administration, will also decrease sleep-related LH pulse frequency in adult women (26, 29, 30) and adolescent girls (31, 32), and 7 days of progesterone administration slows sleep-related, free alpha-subunit (GnRH) pulse frequency in estradiol-replaced postmenopausal women (33). It remains unknown whether progesterone has a differential effect on wake vs sleep LH pulse frequency in adult women, either in terms of degree of suppression or rapidity of suppression. Progesterone action at the hypothalamus appears to require the permissive presence of estradiol (34, 35), presumably because estradiol can increase progesterone receptor expression in the hypothalamus (36, 37). Although the minimum estradiol level needed for effective progesterone action is not known, progesterone acutely reduced LH pulse frequency during waking hours in the current study, suggesting that estradiol concentrations of 27.7 ± 19.9 pg/mL (mean ± SD) were sufficient to induce and/or maintain the relevant hypothalamic progesterone receptors in late pubertal girls. It remains possible that effective progesterone action requires higher estradiol concentrations in other groups (e.g., normal adult women, women with PCOS). Our data showed that LH pulse mass increased with progesterone administration regardless of sleep status. This is in keeping with earlier studies: progesterone increases gonadotropin responses to exogenous GnRH (38–40) and augments gonadotropin secretion in estradiol-pretreated women (14, 16, 41). Although mean LH was much higher with progesterone compared with placebo in three subjects, progesterone did not have a statistically significant overall effect on mean LH or mean FSH. The current study addresses several limitations of our prior study in early pubertal girls (9). First, a relatively small number of subjects (n = 5) received progesterone in our earlier study, and they were compared with a separate group who did not receive progesterone (n = 13). The current study involved 11 subjects evaluated under both progesterone and placebo conditions; the crossover design allowed each girl to serve as her own control, avoiding potential confounding related to between-group differences. Second, it seemed possible (albeit unlikely) that all girls receiving progesterone in our earlier study would have had undetectable wake-associated LH pulse frequency regardless of progesterone administration. In contrast to early pubertal girls, postmenarcheal girls are reliably expected to exhibit daytime LH pulses, and their higher LH pulse amplitudes enhance the reliability of LH pulse detection. Third, blood withdrawal limits in early pubertal girls constrained prior surveillance of daytime LH secretion to 4 hours; in the current study, greater blood volumes permitted frequent sampling during an 8-hour sleep period and for a full 10 hours while awake. Lastly, sleep was not formally evaluated in our earlier study, but we used actigraphy to determine sleep periods in the current study. Regarding potential limitations of this study, we recognize that the use of time blocks as surrogates for wake and sleep periods can be imprecise; however, LH pulse frequency for actigraphy-directed sleep periods correlated very highly with time block-derived estimates for sleep-related LH pulse frequency (rs = 0.977), and the results of actigraphy-directed analysis were virtually identical to the results of time block-directed analysis. Second, we assessed waking LH pulse frequency as the average of two, 4-hour time blocks that were temporally separated, instead of assessment during a continuous 8-hour window (as for sleep-related LH pulse frequency). This decision reflected a desire to minimize the potential impact of time-related confounding: if we had assessed waking pulse frequency from 1500 to 2300 hours or from 0700 to 1500 hours, then sleep-associated pulse frequency would have been associated with progesterone exposure that was 8 hours longer or shorter, respectively. Third, progesterone concentrations achieved in this study were supraphysiologic compared with the normal follicular phase, but this was necessary. Recapitulation of physiological progesterone concentrations would not be expected to alter LH secretion vis-à-vis the placebo condition. Fourth, we studied late pubertal girls to investigate a physiological phenomenon that may be most relevant in early pubertal subjects. However, early pubertal subjects are more difficult to recruit; they less reliably exhibit LH secretion while awake; and their lower body weight translates to more restrictive blood withdrawal limitations, which constrain the duration of LH surveillance and discourage a two-admission crossover design. Lastly, although most subjects had regular menses, we did not directly assess ovulatory status for the cycles immediately preceding overnight admissions, and it remains possible that differences in ovulatory status (e.g., ovulatory vs anovulatory) preceding overnight admissions could have influenced study results. Given the study’s crossover design, where each subject serves as her own control, we assume that any potential confounding would primarily pertain to unequal within-subject study conditions (e.g., if a subject’s first study admission was preceded by an ovulatory cycle, and her second study admission was preceded by an anovulatory cycle). However, randomization of treatment order would be expected to mitigate potential bias in this regard: any bias would occur in one direction in roughly one-half of the subjects (e.g., exaggerating or masking the apparent effect of progesterone) and in the opposite direction in roughly one-half of the subjects. Thus, whereas this putative confounder could have contributed to study outcome variability, it seems unlikely that it could have systematically biased our results or led to type 1 statistical error. In conclusion, this highly rigorous clinical research study demonstrates that progesterone acutely decreases waking LH pulse frequency to a greater extent than sleep-associated LH frequency in late pubertal girls without hyperandrogenemia. Taken together with our prior study in early-to-midpubertal girls, these data strongly support the hypothesis that the pubertal GnRH pulse generator is differentially sensitive to progesterone negative feedback depending on sleep status. This may represent an important mechanism directing day-night changes of GnRH pulse frequency during normal puberty. Abbreviations: CRU Clinical Research Unit CV coefficient of variation FSH follicle-stimulating hormone GnRH gonadotropin-releasing hormone LC-MS/MS liquid chromatography-tandem mass spectrometry LH luteinizing hormone LM linear-mixed PCOS polycystic ovary syndrome RIA radioimmunoassay rs Spearman’s rank correlation coefficient SD standard deviation SHBG sex hormone-binding globulin UVA University of Virginia. Acknowledgments We gratefully acknowledge Anne Gabel, Katherine Ehrlich, Deborah M. Sanderson, Melissa Gilrain, Amy Anderson, and Eleanor Hutchens for subject recruitment, study scheduling, and assistance with data management. We also extend our gratitude to the nurses and staff of the Clinical Research Unit at University of Virginia for implementation of the sampling protocols and to Dan Haisenleder and the Center for Research in Reproduction Ligand Core Laboratory for performance of assays. Financial Support: This work was supported by US National Institutes of Health (NIH) Grant R01 HD058671 (to C.R.M. and J.T.P.); Eunice Kennedy Shriver National Institute of Child Health and Human Development/NIH through Cooperative Agreement P50 HD28934 as part of the National Centers for Translational Research in Reproduction and Infertility (to C.R.M., J.C.M., C.M.B.S., and J.T.P.); and NIH Grants F32 HD088047 (to S.H.K.), F32 HD091951 (to J.A.L.), F32 HD078088 (to R.B.), F32 HD066855 (to J.S.C.), and K23 HD070854 (to C.M.B.S.). Clinical Trial Information: ClinicalTrials.gov no. NCT00929006 (registered 26 June 2009). Author Contributions: S.H.K. was the primary individual performing data collection, organization, and analysis, and she was the primary author of manuscript. J.A.L., R.B., and J.S.C. contributed to study design and implementation, including subject recruitment, oversight of CRU admissions, and data collection/analysis, and they provided editorial guidance during manuscript preparation. J.T.P. established all statistical plans before study initiation, provided the randomization schedule to the Investigational Pharmacy, and performed all primary and secondary statistical analyses described herein, and he provided editorial guidance during manuscript preparation. C.M.B.S. and J.C.M. provided input into study design and data interpretation, and they provided editorial guidance during manuscript preparation. C.R.M. was the principal investigator of the study and was responsible for study design and oversight. He supervised data collection and organization, performed all pulse detection analyses, directed all data analysis, and materially participated in manuscript preparation. Disclosure Summary: The authors have nothing to disclose. References 1. McCartney CR. Maturation of sleep-wake gonadotrophin-releasing hormone secretion across puberty in girls: potential mechanisms and relevance to the pathogenesis of polycystic ovary syndrome. J Neuroendocrinol . 2010; 22( 7): 701– 709. Google Scholar PubMed  2. Boyar R, Finkelstein J, Roffwarg H, Kapen S, Weitzman E, Hellman L. Synchronization of augmented luteinizing hormone secretion with sleep during puberty. N Engl J Med . 1972; 287( 12): 582– 586. Google Scholar CrossRef Search ADS PubMed  3. Apter D, Bützow TL, Laughlin GA, Yen SS. Gonadotropin-releasing hormone pulse generator activity during pubertal transition in girls: pulsatile and diurnal patterns of circulating gonadotropins. J Clin Endocrinol Metab . 1993; 76( 4): 940– 949. Google Scholar PubMed  4. Shaw ND, Butler JP, McKinney SM, Nelson SA, Ellenbogen JM, Hall JE. Insights into puberty: the relationship between sleep stages and pulsatile LH secretion. J Clin Endocrinol Metab . 2012; 97( 11): E2055– E2062. Google Scholar CrossRef Search ADS PubMed  5. Burger LL, Haisenleder DJ, Dalkin AC, Marshall JC. Regulation of gonadotropin subunit gene transcription. J Mol Endocrinol . 2004; 33( 3): 559– 584. Google Scholar CrossRef Search ADS PubMed  6. Thompson IR, Kaiser UB. GnRH pulse frequency-dependent differential regulation of LH and FSH gene expression. Mol Cell Endocrinol . 2014; 385( 1-2): 28– 35. Google Scholar CrossRef Search ADS PubMed  7. Burt Solorzano CM, Beller JP, Abshire MY, Collins JS, McCartney CR, Marshall JC. Neuroendocrine dysfunction in polycystic ovary syndrome. Steroids . 2012; 77( 4): 332– 337. Google Scholar CrossRef Search ADS PubMed  8. McCartney CR, Blank SK, Prendergast KA, Chhabra S, Eagleson CA, Helm KD, Yoo R, Chang RJ, Foster CM, Caprio S, Marshall JC. Obesity and sex steroid changes across puberty: evidence for marked hyperandrogenemia in pre- and early pubertal obese girls. J Clin Endocrinol Metab . 2007; 92( 2): 430– 436. Google Scholar CrossRef Search ADS PubMed  9. Collins JS, Marshall JC, McCartney CR. Differential sleep-wake sensitivity of gonadotropin-releasing hormone secretion to progesterone inhibition in early pubertal girls. Neuroendocrinology . 2012; 96( 3): 222– 227. Google Scholar CrossRef Search ADS PubMed  10. McCartney CR, Prendergast KA, Blank SK, Helm KD, Chhabra S, Marshall JC. Maturation of luteinizing hormone (gonadotropin-releasing hormone) secretion across puberty: evidence for altered regulation in obese peripubertal girls. J Clin Endocrinol Metab . 2009; 94( 1): 56– 66. Google Scholar CrossRef Search ADS PubMed  11. Wierman ME, Auchus RJ, Haisenleder DJ, Hall JE, Handelsman D, Hankinson S, Rosner W, Singh RJ, Sluss PM, Stanczyk FZ. Editorial: The new instructions to authors for the reporting of steroid hormone measurements. J Clin Endocrinol Metab . 2014; 99( 12): 4375. Google Scholar CrossRef Search ADS PubMed  12. Bhasin S, Pencina M, Jasuja GK, Travison TG, Coviello A, Orwoll E, Wang PY, Nielson C, Wu F, Tajar A, Labrie F, Vesper H, Zhang A, Ulloor J, Singh R, D’Agostino R, Vasan RS. Reference ranges for testosterone in men generated using liquid chromatography tandem mass spectrometry in a community-based sample of healthy nonobese young men in the Framingham Heart Study and applied to three geographically distinct cohorts. J Clin Endocrinol Metab . 2011; 96( 8): 2430– 2439. Google Scholar CrossRef Search ADS PubMed  13. Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab . 1999; 84( 10): 3666– 3672. Google Scholar CrossRef Search ADS PubMed  14. Hutchens EG, Ramsey KA, Howard LC, Abshire MY, Patrie JT, McCartney CR. Progesterone has rapid positive feedback actions on LH release but fails to reduce LH pulse frequency within 12 h in estradiol-pretreated women. Physiol Rep . 2016; 4( 16): e12891. Google Scholar CrossRef Search ADS PubMed  15. Johnson ML, Pipes L, Veldhuis PP, Farhy LS, Boyd DG, Evans WS. AutoDecon, a deconvolution algorithm for identification and characterization of luteinizing hormone secretory bursts: description and validation using synthetic data. Anal Biochem . 2008; 381( 1): 8– 17. Google Scholar CrossRef Search ADS PubMed  16. McCartney CR, Blank SK, Marshall JC. Progesterone acutely increases LH pulse amplitude but does not acutely influence nocturnal LH pulse frequency slowing during the late follicular phase in women. Am J Physiol Endocrinol Metab . 2007; 292( 3): E900– E906. Google Scholar CrossRef Search ADS PubMed  17. McCartney CR, Marshall JC. Neuroendocrinology of reproduction. In: Strauss JF, Barbieri RL, eds. Yen & Jaffe's Reproductive Endocrinology: Physiology, Pathophysiology, and Clinical Management . 7th ed. Philadelphia, PA: Elsevier Saunders, 2014: 3– 26 18. Hall JE. Neuroendocrine control of the menstrual cycle. In: Strauss JF, Barbieri RL, eds. Yen & Jaffe's Reproductive Endocrinology: Physiology, Pathophysiology, and Clinical Management . 7th ed. Philadelphia, PA: Elsevier Saunders, 2014: 141– 156 19. Cemeroglu AP, Foster CM, Warner R, Kletter GB, Marshall JC, Kelch RP. Comparison of the neuroendocrine control of pubertal maturation in girls and boys with spontaneous puberty and in hypogonadal girls. J Clin Endocrinol Metab . 1996; 81( 12): 4352– 4357. Google Scholar PubMed  20. Loucks AB, Heath EM. Dietary restriction reduces luteinizing hormone (LH) pulse frequency during waking hours and increases LH pulse amplitude during sleep in young menstruating women. J Clin Endocrinol Metab . 1994; 78( 4): 910– 915. Google Scholar PubMed  21. Loucks AB, Verdun M, Heath EM. Low energy availability, not stress of exercise, alters LH pulsatility in exercising women. J Appl Physiol (1985) . 1998; 84( 1): 37– 46. Google Scholar CrossRef Search ADS PubMed  22. Ankarberg C, Norjavaara E. Diurnal rhythm of testosterone secretion before and throughout puberty in healthy girls: correlation with 17beta-estradiol and dehydroepiandrosterone sulfate. J Clin Endocrinol Metab . 1999; 84( 3): 975– 984. Google Scholar PubMed  23. Mitamura R, Yano K, Suzuki N, Ito Y, Makita Y, Okuno A. Diurnal rhythms of luteinizing hormone, follicle-stimulating hormone, testosterone, and estradiol secretion before the onset of female puberty in short children. J Clin Endocrinol Metab . 2000; 85( 3): 1074– 1080. Google Scholar PubMed  24. Eagleson CA, Gingrich MB, Pastor CL, Arora TK, Burt CM, Evans WS, Marshall JC. Polycystic ovarian syndrome: evidence that flutamide restores sensitivity of the gonadotropin-releasing hormone pulse generator to inhibition by estradiol and progesterone. J Clin Endocrinol Metab . 2000; 85( 11): 4047– 4052. Google Scholar PubMed  25. Pielecka J, Quaynor SD, Moenter SM. Androgens increase gonadotropin-releasing hormone neuron firing activity in females and interfere with progesterone negative feedback. Endocrinology . 2006; 147( 3): 1474– 1479. Google Scholar CrossRef Search ADS PubMed  26. Pastor CL, Griffin-Korf ML, Aloi JA, Evans WS, Marshall JC. Polycystic ovary syndrome: evidence for reduced sensitivity of the gonadotropin-releasing hormone pulse generator to inhibition by estradiol and progesterone. J Clin Endocrinol Metab . 1998; 83( 2): 582– 590. Google Scholar PubMed  27. Filicori M, Santoro N, Merriam GR, Crowley WF Jr. Characterization of the physiological pattern of episodic gonadotropin secretion throughout the human menstrual cycle. J Clin Endocrinol Metab . 1986; 62( 6): 1136– 1144. Google Scholar CrossRef Search ADS PubMed  28. Rossmanith WG, Liu CH, Laughlin GA, Mortola JF, Suh BY, Yen SS. Relative changes in LH pulsatility during the menstrual cycle: using data from hypogonadal women as a reference point. Clin Endocrinol (Oxf) . 1990; 32( 5): 647– 660. Google Scholar CrossRef Search ADS PubMed  29. Soules MR, Steiner RA, Clifton DK, Cohen NL, Aksel S, Bremner WJ. Progesterone modulation of pulsatile luteinizing hormone secretion in normal women. J Clin Endocrinol Metab . 1984; 58( 2): 378– 383. Google Scholar CrossRef Search ADS PubMed  30. McCartney CR, Gingrich MB, Hu Y, Evans WS, Marshall JC. Hypothalamic regulation of cyclic ovulation: evidence that the increase in gonadotropin-releasing hormone pulse frequency during the follicular phase reflects the gradual loss of the restraining effects of progesterone. J Clin Endocrinol Metab . 2002; 87( 5): 2194– 2200. Google Scholar CrossRef Search ADS PubMed  31. Chhabra S, McCartney CR, Yoo RY, Eagleson CA, Chang RJ, Marshall JC. Progesterone inhibition of the hypothalamic gonadotropin-releasing hormone pulse generator: evidence for varied effects in hyperandrogenemic adolescent girls. J Clin Endocrinol Metab . 2005; 90( 5): 2810– 2815. Google Scholar CrossRef Search ADS PubMed  32. Blank SK, McCartney CR, Chhabra S, Helm KD, Eagleson CA, Chang RJ, Marshall JC. Modulation of gonadotropin-releasing hormone pulse generator sensitivity to progesterone inhibition in hyperandrogenic adolescent girls--implications for regulation of pubertal maturation. J Clin Endocrinol Metab . 2009; 94( 7): 2360– 2366. Google Scholar CrossRef Search ADS PubMed  33. Gill S, Lavoie HB, Bo-Abbas Y, Hall JE. Negative feedback effects of gonadal steroids are preserved with aging in postmenopausal women. J Clin Endocrinol Metab . 2002; 87( 5): 2297– 2302. Google Scholar CrossRef Search ADS PubMed  34. Karsch FJ, Weick RF, Hotchkiss J, Dierschke DJ, Knobil E. An analysis of the negative feedback control of gonadotropin secretion utilizing chronic implantation of ovarian steroids in ovariectomized rhesus monkeys. Endocrinology . 1973; 93( 2): 478– 486. Google Scholar CrossRef Search ADS PubMed  35. Nippoldt TB, Reame NE, Kelch RP, Marshall JC. The roles of estradiol and progesterone in decreasing luteinizing hormone pulse frequency in the luteal phase of the menstrual cycle. J Clin Endocrinol Metab . 1989; 69( 1): 67– 76. Google Scholar CrossRef Search ADS PubMed  36. MacLusky NJ, McEwen BS. Oestrogen modulates progestin receptor concentrations in some rat brain regions but not in others. Nature . 1978; 274( 5668): 276– 278. Google Scholar CrossRef Search ADS PubMed  37. Romano GJ, Krust A, Pfaff DW. Expression and estrogen regulation of progesterone receptor mRNA in neurons of the mediobasal hypothalamus: an in situ hybridization study. Mol Endocrinol . 1989; 3( 8): 1295– 1300. Google Scholar CrossRef Search ADS PubMed  38. Chang RJ, Jaffe RB. Progesterone effects on gonadotropin release in women pretreated with estradiol. J Clin Endocrinol Metab . 1978; 47( 1): 119– 125. Google Scholar CrossRef Search ADS PubMed  39. Nippoldt TB, Khoury S, Barkan A, Kelch RP, Marshall JC. Gonadotrophin responses to GnRH pulses in hypogonadotrophic hypogonadism: LH responsiveness is maintained in the presence of luteal phase concentrations of oestrogen and progesterone. Clin Endocrinol (Oxf) . 1987; 26( 3): 293– 301. Google Scholar CrossRef Search ADS PubMed  40. Couzinet B, Brailly S, Bouchard P, Schaison G. Progesterone stimulates luteinizing hormone secretion by acting directly on the pituitary. J Clin Endocrinol Metab . 1992; 74( 2): 374– 378. Google Scholar PubMed  41. Liu JH, Yen SS. Induction of midcycle gonadotropin surge by ovarian steroids in women: a critical evaluation. J Clin Endocrinol Metab . 1983; 57( 4): 797– 802. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

Progesterone-Mediated Inhibition of the GnRH Pulse Generator: Differential Sensitivity as a Function of Sleep Status

Loading next page...
 
/lp/ou_press/progesterone-mediated-inhibition-of-the-gnrh-pulse-generator-CghfxHOwxT
Publisher
Endocrine Society
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2017-02299
Publisher site
See Article on Publisher Site

Abstract

Abstract Context During normal, early puberty, luteinizing hormone (LH) pulse frequency is low while awake but increases during sleep. Mechanisms underlying such changes are unclear, but a small study in early pubertal girls suggested that differential wake-sleep sensitivity to progesterone negative feedback plays a role. Objective To test the hypothesis that progesterone acutely reduces waking LH pulse frequency more than sleep-associated pulse frequency in late pubertal girls. Design Randomized, placebo-controlled, double-blinded crossover study. Setting Academic clinical research unit. Participants Eleven normal, postmenarcheal girls, ages 12 to 15 years. Intervention Subjects completed two 18-hour admissions in separate menstrual cycles (cycle days 6 to 11). Frequent blood sampling for LH assessment was performed at 1800 to 1200 hours; sleep was encouraged at 2300 to 0700 hours. Either oral micronized progesterone (0.8 mg/kg/dose) or placebo was given at 0700, 1500, 2300, and 0700 hours, before and during the first admission. A second admission, performed at least 2 months later, was identical to the first except that placebo was exchanged for progesterone or vice versa (treatment crossover). Main Outcome Measures LH pulse frequency during waking and sleeping hours. Results Progesterone reduced waking LH pulse frequency by 26% (P = 0.019), with no change observed during sleep (P = 0.314). The interaction between treatment condition (progesterone vs placebo) and sleep status (wake vs sleep) was highly significant (P = 0.007). Conclusions In late pubertal girls, progesterone acutely reduced waking LH pulse frequency more than sleep-associated pulse frequency. Differential wake-sleep sensitivity to progesterone negative feedback may direct sleep-wake LH pulse frequency changes across puberty. The hypothalamic-pituitary-gonadal axis is quiescent during childhood, and nocturnal amplification of luteinizing hormone [LH; mirroring gonadotropin-releasing hormone (GnRH)] pulse frequency and amplitude marks the beginning of puberty (1). Mechanisms underlying these marked sleep-wake differences in LH secretion remain unclear, but nocturnal amplification of LH release during puberty is clearly tied to sleep: it generally occurs within 1 hour of sleep and follows sleep reversal (2–4). As high and low GnRH pulse frequencies favor LH and follicle-stimulating hormone (FSH) production, respectively (1, 5, 6), these patterns may be critical for normal early pubertal development, and disruptions of such may be relevant to polycystic ovary syndrome (PCOS), a disorder characterized by abnormal gonadotropin secretion that frequently manifests around the time of puberty (7). Our group has demonstrated that in early pubertal girls, progesterone concentrations increase in the early morning, and we hypothesized that this may rapidly reduce morning GnRH pulse frequency, contributing to day-night differences in LH release (8). To test this notion, we assessed LH pulse secretion in early-to-midpubertal girls who did and did not receive exogenous progesterone in the evening, hypothesizing that progesterone would slow nocturnal LH pulse frequency. Whereas progesterone appeared to abolish late-evening LH pulse frequency (while awake), it did not alter nocturnal (sleep-related) LH pulse frequency (9). These data suggested that the pubertal GnRH pulse generator is differentially sensitive to progesterone inhibition based on sleep-wake status. However, this study involved a relatively small number of subjects receiving progesterone and limited surveillance of daytime LH secretion, and as some early pubertal girls exhibit little daytime LH secretion, it remained possible that some girls in the progesterone group would have exhibited no daytime LH secretion regardless. To test more rigorously the hypothesis that exogenous progesterone acutely reduces waking LH pulse frequency to a greater extent than sleep-associated LH pulse frequency in pubertal girls, we performed a randomized, placebo-controlled, double-blinded crossover study in late pubertal girls. Subjects and Methods Eleven healthy, late pubertal girls without hyperandrogenism were studied at the University of Virginia (UVA). Subjects were 12 to 15 years old, and all were postmenarcheal. Tanner breast stage was 3.9 ± 1.0 [mean ± standard deviation (SD)]. Eight subjects reported regular menstrual cycles. One subject, who was 1 year postmenarcheal, described menses every 2 months, and two subjects, who were postmenarcheal by <6 months, had not yet established a clear pattern. None had evidence for hormonal abnormalities (e.g., hyperandrogenism) or premature adrenarche. None had used medications known to affect the reproductive system, glucose metabolism, lipid metabolism, or blood pressure within 90 days of the study. Subject characteristics are shown in Table 1. Table 1. Subject Characteristics   n  Means  SD  Median  Range  Age, y  11  14.2  1.1  14.3  12.0–15.9  Bone age, y  10  15.0  1.3  15  13.5–17.0  Tanner breast stage  11  3.9  1.0  4  2–5  Years postmenarche  11  1.6  1.1  1  0–3  Cycle length, d  9a  33.6  10.8  30  24–60  BMI, kg/m2  11  27.5  11.5  23.6  18.6–57.4  BMI-for-age percentile  11  77.5  22.4  88  37–99  BMI z-score  11  1.09  1.05  1.15  −0.33 to 2.82  Waist circumference, cm  11  83.1  27.3  70.2  62.0–155.3  Hip circumference, cm  11  101.9  21.4  95.2  82.5–158.6  Waist-to-hip ratio  11  0.8  0.09  0.75  0.73–0.98  Body fat percentage  11  32.3  11.4  30.8  16.7–53.5  Total testosterone, ng/dLb  11  19.1  6.1  19  7.5–28.0  SHBG, nmol/Lb  11  42.4  16.3  46.7  18.7–76.7  Free testosterone, pg/mLb  11  3.3  1.5  3.0  1.5–5.4  Androstenedione, ng/dLb  11  176  96  157  68–376  DHEA-S, µg/dLb  11  104  59  83  50–247  Estradiol, pg/mLb  11  27.1  15.5  21.7  12.5–63.2  Fasting insulin, µIU/mLb  11  14.2  7.1  12.0  4.2–25.1  Fasting glucose, mg/dLb  10  94.3  7.2  94  86–111c  Hemoglobin A1c, %  10  5.2  0.3  5.2  4.8–5.6  IGF-1, ng/mLb  11  331  114  314  194–531    n  Means  SD  Median  Range  Age, y  11  14.2  1.1  14.3  12.0–15.9  Bone age, y  10  15.0  1.3  15  13.5–17.0  Tanner breast stage  11  3.9  1.0  4  2–5  Years postmenarche  11  1.6  1.1  1  0–3  Cycle length, d  9a  33.6  10.8  30  24–60  BMI, kg/m2  11  27.5  11.5  23.6  18.6–57.4  BMI-for-age percentile  11  77.5  22.4  88  37–99  BMI z-score  11  1.09  1.05  1.15  −0.33 to 2.82  Waist circumference, cm  11  83.1  27.3  70.2  62.0–155.3  Hip circumference, cm  11  101.9  21.4  95.2  82.5–158.6  Waist-to-hip ratio  11  0.8  0.09  0.75  0.73–0.98  Body fat percentage  11  32.3  11.4  30.8  16.7–53.5  Total testosterone, ng/dLb  11  19.1  6.1  19  7.5–28.0  SHBG, nmol/Lb  11  42.4  16.3  46.7  18.7–76.7  Free testosterone, pg/mLb  11  3.3  1.5  3.0  1.5–5.4  Androstenedione, ng/dLb  11  176  96  157  68–376  DHEA-S, µg/dLb  11  104  59  83  50–247  Estradiol, pg/mLb  11  27.1  15.5  21.7  12.5–63.2  Fasting insulin, µIU/mLb  11  14.2  7.1  12.0  4.2–25.1  Fasting glucose, mg/dLb  10  94.3  7.2  94  86–111c  Hemoglobin A1c, %  10  5.2  0.3  5.2  4.8–5.6  IGF-1, ng/mLb  11  331  114  314  194–531  To convert conventional to international system of units: total testosterone (ng/dL) × 3.467 (nmol/L); sex hormone-binding globulin (SHBG; µg/mL) × 8.896 (nmol/L); free testosterone (pg/mL) × 3.467 (pmol/L); androstenedione (ng/dL) × 0.131 (nmol/L); dehydroepiandrosterone sulfate (DHEA-S; µg/dL) × 27.211 (nmol/L); estradiol (pg/mL) × 3.671 (pmol/L); insulin (µIU/mL) × 7.175 (pmol/L); glucose (mg/dL) × 0.0555 (mmol/L); and insulinlike growth factor 1 (IGF-1; ng/mL) × 0.131 (nmol/L). Abbreviation: BMI, body mass index. a Summary data for cycle length does not include data for two girls who were <6 months postmenarcheal at the start of study. b Reported values were obtained from the placebo admission. Differences between placebo and progesterone admissions were assessed with Wilcoxon rank sum tests for androstenedione, DHEA-S, insulin, glucose, and IGF-1; all P values were >0.7. c Only one subject had a fasting glucose >100 during the placebo admission, but her fasting glucose levels from screening and the progesterone admission were 86 and 96, respectively. View Large Study procedures The Institutional Review Board at UVA approved all study procedures, which were in accordance with the ethical standards of the Helsinki Declaration of 1975, as revised in 2008. The study was registered with ClinicalTrials.gov (identifier NCT00929006). Informed assent and consent were obtained from all subjects and their custodial parents, respectively. Subjects underwent a detailed, physician-obtained medical history and physical examination, which included determination of pubertal stage (Tanner scale for breast development). As previously described (10), fasting blood tests were drawn at 0800 to 0900 hours to ensure good general health and to exclude hormonal abnormalities. All screening blood test results were unremarkable for each subject. Measures of adiposity included waist circumference and hip circumference, and BOD POD® was used to measure percent body fat. Plain x-ray of the left hand and wrist was obtained to assess bone age. Subjects underwent two separate admissions for frequent blood sampling (LH pulse analysis) in the UVA Clinical Research Unit (CRU; Fig. 1). All admissions occurred in the mid-to-late follicular phase (cycle days 6 to 11). Red blood cell counts, β-human chorionic gonadotropin, and plasma progesterone levels were checked 1 to 3 days before each overnight admission to exclude anemia, pregnancy, and the luteal phase, respectively. Figure 1. View largeDownload slide Study schematic. Vertical arrows indicate the times when either oral micronized progesterone (P4) or placebo (PBO) was given. PBO was exchanged for P4 or vice versa for the second admission (treatment crossover). Horizontal arrows indicate wake (1900 to 2300 and 0700 to 1100 hours) and sleep (2300 to 0300 and 0300 to 0700 hours) blocks. Figure 1. View largeDownload slide Study schematic. Vertical arrows indicate the times when either oral micronized progesterone (P4) or placebo (PBO) was given. PBO was exchanged for P4 or vice versa for the second admission (treatment crossover). Horizontal arrows indicate wake (1900 to 2300 and 0700 to 1100 hours) and sleep (2300 to 0300 and 0300 to 0700 hours) blocks. Subjects were randomized to receive either progesterone or placebo doses before (i.e., at 0700 and 1500 hours) and during (i.e., at 2300 and 0700 hours) the first overnight admission. Progesterone was administered as a cherry-flavored micronized progesterone suspension compounded by the UVA Investigational Pharmacy (US Food and Drug Administration investigational new drug 64,126); the placebo solution was identical, except that micronized progesterone was omitted. Each progesterone dose was 0.8 mg/kg body weight, which generally produces a mean progesterone level of 3 to 8 ng/mL when given three times daily. Subjects, CRU staff, and investigators were blinded to treatment condition. Subjects were admitted to the CRU at 1600 hours. Beginning at 1800 hours, blood was obtained through an indwelling, intravenous forearm catheter over an 18-hour period (1800 to 1200 hours): LH every 10 minutes; progesterone every 30 minutes; FSH, estradiol, and total testosterone every hour. Additional 1-mL samples taken at 1800, 0000, 0600, and 1200 hours were pooled for measurement of total testosterone by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Lights were extinguished at 2300 hours. Sleep was encouraged from 2300 to 0700 hours, but sleep was otherwise disallowed. Sleep periods were formally evaluated using wrist actigraphy (Motionlogger Basic-L; Ambulatory Monitoring, Ardsley, NY). Subjects fasted from 2300 to 0700 hours; at 0700 hours, an additional 4 mL was drawn for measurement of insulin, glucose, sex hormone-binding globulin (SHBG), dehydroepiandrosterone-sulfate, insulinlike growth factor 1, and androstenedione. Subjects were discharged after the final blood draw at 1200 hours. At least 2 months after the first admission, subjects returned for a second overnight admission identical to the first, except that exogenous progesterone was exchanged for placebo, or vice versa, in accordance with the crossover design. We enforced a 2-month minimum between admissions to allow adequate restoration of the subject’s red blood cell population. Hormonal measurements UVA’s Center for Research in Reproduction Ligand Assay and Analysis Core performed all hormone assays except for total testosterone by LC-MS/MS. Hormone concentrations measured below assay sensitivity were assigned the value of the assay’s sensitivity. LH was measured by chemiluminescence [Diagnostic Products, Los Angeles, CA; sensitivity, 0.1 IU/L; intra-assay coefficient of variation (CV) ≤3.3%; interassay CV ≤6.6%]. Each subject’s LH samples were analyzed in duplicate in the same assay. Other assay characteristics are described in detail in the Supplemental Materials. Before January 2015, sex steroids were measured by radioimmunoassay (RIA), as previously described (10), but in 2014, most steroid RIA kits used by the Center for Research in Reproduction were discontinued by the manufacturer and replaced by other immunoassays. Evaluations of replacement methods were conducted according to recommendations of the Endocrine Society Sex Steroid Assays Reporting Task Force (11). Each replacement method was carefully correlated to Diagnostic Products RIA, and RIA-equivalent values are presented herein. Total testosterone by LC-MS/MS was performed by the Mayo Clinic Laboratory [sensitivity 7 ng/dL; intra-assay CV ≤6.0%; interassay CVs 7.9% to 15.8% at 12 to 48.6 ng/dL (12)]. Total testosterone measurements by immunoassay and LC-MS/MS were highly concordant (Supplemental Materials). However, as the same LC-MS/MS testosterone assay was performed for all subjects in this study—in contrast to the use of different testosterone immunoassay methodologies over time—we elected to use LC-MS/MS results for formal analyses. Free testosterone was calculated from total testosterone and SHBG (13). Data analysis Data analyses related to pulsatile LH secretion was performed by a single investigator (C.R.M.) while blinded to treatment condition. As previously described (14), a computerized data reduction protocol (StdCurve) was used to provide a variance model for experimental measurement error for each LH concentration time series. LH pulses were then identified using the computerized pulse detection algorithm AutoDecon, a fully automated (i.e., nonsubjective), multiparameter deconvolution program (15). To limit false positives, only the following AutoDecon-identified pulses were included in subsequent analyses: (1) pulses with at least two peak values that were at least 10% higher than the preceding nadir or (2) pulses with at least one peak value that was at least 20% higher than the preceding nadir. Data derived from AutoDecon included the temporal locations of LH pulses, which were used to calculate LH pulse frequency, and LH pulse mass, an estimate of the amount of LH secreted by the pituitary during each pulse (a correlate of LH pulse amplitude). With the use of previously described methods (16), average LH interpulse intervals were calculated during four time blocks: 1900 to 2300, 2300 to 0300, 0300 to 0700, and 0700 to 1100 hours. For each time block, LH pulse frequency (expressed as pulses per hour) was then calculated by dividing 60 by the average interpulse interval. Our method of LH pulse-frequency determination for discrete time blocks is described in detail in Supplemental Materials. Wake LH pulse frequency was defined as the average of 1900 to 2300- and 0700 to 1100-hour blocks, and sleep LH pulse frequency was defined as the average of 2300 to 0300- and 0300 to 0700-hour blocks. LH pulse frequency was similarly calculated for actigraphy-determined sleep periods (between the first and last sleep epochs) and wake periods [weighted average of (1) frequency before first sleep and (2) frequency after last sleep]. Statistical analyses As the a priori primary statistical analysis, LH pulse frequency during wake and sleep periods of the placebo and progesterone admissions was analyzed in accordance with a two-period crossover design paradigm by way of linear-mixed (LM) models. The LM model was specified so that the variability in LH pulse frequency related to sequence order (progesterone vs placebo given first) and crossover period (first CRU admission vs second CRU admission) could be extracted from the variability in LH pulse frequency related to sleep status (sleep vs wake), treatment (progesterone vs placebo), and the sleep status by treatment interaction. With regard to hypothesis testing, the LM model covariance matrix was specified in the unstructured form, and linear contrasts of the least-square means were used to test both (1) if progesterone-induced changes in LH pulse frequency under the sleep and wake conditions were equal to zero and (2) if progesterone-induced changes in LH pulse frequency were the same under sleep and wake conditions. As preplanned secondary statistical analyses, LH pulse mass, mean LH, and mean FSH of the wake and sleep periods of the placebo and progesterone admissions were analyzed using LM models, as described previously. Changes in mean progesterone, estradiol, total testosterone, and free testosterone concentrations between admissions were analyzed via paired Student’s t tests. As post hoc analyses, we performed Wilcoxon signed-rank tests to assess for progesterone-related LH pulse-frequency differences (progesterone vs placebo admission) for each 4-hour time block (i.e., 1900 to 2300, 2300 to 0300, 0300 to 0700, and 0700 to 1100 hours). Spearman’s rank correlation coefficient (rs) was used to assess whether progesterone-induced changes in waking LH pulse frequency were associated with baseline free testosterone concentration. SAS version 9.4 (SAS Institute, Cary, NC) was used for all statistical analyses. A two-sided P ≤ 0.05 decision rule was used as the null hypothesis rejection criterion for all statistical tests. Data are presented as means ± standard error of the mean unless indicated otherwise. Subject-level data used for analyses are provided in Supplemental Materials. Results Sex steroid concentrations and actigraphy data Mean progesterone concentrations were higher during progesterone admissions compared with placebo admissions (6.5 ± 1.3 vs 0.4 ± 0.03 ng/mL; P = 0.001; Fig. 2). Mean estradiol was similar between admissions (27.7 ± 6.0 and 27.1 ± 4.7 pg/mL, respectively). Total testosterone by LC-MS/MS was lower during progesterone admissions compared with placebo admissions (15.3 ± 2.1 vs 19.1 ± 1.8 ng/dL; P = 0.047). SHBG concentrations were similar between progesterone and placebo admissions (40.8 ± 6.8 vs 42.4 ± 4.9 nmol/L, respectively; P = 0.757), as were calculated free testosterone concentrations using LC-MS/MS (2.7 ± 0.5 vs 3.3 ± 0.4 pg/mL, respectively; P = 0.219). Figure 2. View largeDownload slide Sex steroid changes between progesterone (P4) and placebo (PBO) conditions. Data points indicate the change in (a) mean P4, (b) estradiol, (c) total testosterone, and (d) free testosterone concentrations from the PBO to the P4 condition. Change in hormone concentration in individual subjects (○); these data are also summarized using box-and-whisker plots, which show median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). P values relate to the mean changes between the treatment conditions. E2, estradiol; T, testosterone. Figure 2. View largeDownload slide Sex steroid changes between progesterone (P4) and placebo (PBO) conditions. Data points indicate the change in (a) mean P4, (b) estradiol, (c) total testosterone, and (d) free testosterone concentrations from the PBO to the P4 condition. Change in hormone concentration in individual subjects (○); these data are also summarized using box-and-whisker plots, which show median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). P values relate to the mean changes between the treatment conditions. E2, estradiol; T, testosterone. Per wrist actigraphy, sleep duration (time from first sleep epoch to last sleep epoch) was 7.7 ± 0.7 and 7.7 ± 0.5 hours (means ± SD) during progesterone and placebo admissions, respectively; sleep latency (i.e., time from lights out to first sleep) was 19 ± 43 and 19 ± 33 minutes, respectively; and sleep efficiency (i.e., percentage of sleep duration occupied by sleep epochs) was 89 ± 8 and 92 ± 3%, respectively. Effect of progesterone on wake vs sleep-related LH pulse frequency Average waking LH pulse frequency was 26% lower during progesterone admissions compared with placebo admissions (0.74 ± 0.03 vs 1.00 ± 0.02 pulses/h; P = 0.019; Fig. 3). Average sleep-related LH pulse frequencies were 0.55 ± 0.02 and 0.64 ± 0.02 pulses per hour during progesterone and placebo admissions, respectively (P = 0.314). The interaction between treatment condition and sleep status was highly significant (P = 0.007). Whereas sleep status was an important predictor of LH pulse frequency (P < 0.001), sequence order and crossover period were not (P > 0.6 for both). Figure 3. View largeDownload slide LH pulse frequency as a function of sleep status and treatment condition. (a) Individual average LH pulse frequency under the placebo (PBO) and progesterone (P4) conditions during both (left) wake periods and (right) sleep periods. An individual subject’s data are represented by connected ○. Note that wake LH pulse frequency was lower with P4 than with PBO in most (10 of 11) subjects, whereas sleep-related LH pulse frequency was lower with P4 in less than one-half (five of 11) of subjects. (b) Mean LH pulse frequency (denoted by ●) in each treatment condition during (left) wake and (right) sleep blocks, with vertical lines denoting 95% confidence intervals. P values in each column relate to the change in LH pulse frequency between treatment conditions. P values relating to the interaction between treatment (PBO vs P4) and sleep status (wake vs sleep) are also shown. Figure 3. View largeDownload slide LH pulse frequency as a function of sleep status and treatment condition. (a) Individual average LH pulse frequency under the placebo (PBO) and progesterone (P4) conditions during both (left) wake periods and (right) sleep periods. An individual subject’s data are represented by connected ○. Note that wake LH pulse frequency was lower with P4 than with PBO in most (10 of 11) subjects, whereas sleep-related LH pulse frequency was lower with P4 in less than one-half (five of 11) of subjects. (b) Mean LH pulse frequency (denoted by ●) in each treatment condition during (left) wake and (right) sleep blocks, with vertical lines denoting 95% confidence intervals. P values in each column relate to the change in LH pulse frequency between treatment conditions. P values relating to the interaction between treatment (PBO vs P4) and sleep status (wake vs sleep) are also shown. Notably, LH pulse frequency for actigraphy-directed sleep periods correlated very highly with average LH pulse frequency from 2300 to 0700 hours (rs = 0.977, P < 0.0001), and results were essentially identical when actigraphy-estimated sleep and wake periods were used for analysis: LH pulse frequency during wake periods was lower with progesterone compared with placebo (P = 0.021), with no difference during sleep periods (P = 0.417), and the interaction between treatment condition and sleep status was highly significant (P = 0.007). Effect of progesterone on LH pulse mass, mean LH, and mean FSH LH pulse mass was higher with progesterone compared with placebo during both wake periods (5.32 ± 0.23 vs 2.44 ± 0.13 IU/L; P = 0.006) and sleep periods [7.47 ± 0.30 vs 3.75 ± 0.16 IU; P = 0.016; Fig. 4(a) and 4(b)]. Important predictors of LH pulse mass included sleep status (P = 0.001) but not sequence order or crossover period, and there was no interaction between treatment condition and sleep status. Figure 4. View largeDownload slide Secondary outcomes: LH pulse mass, mean LH, and mean FSH as a function of sleep status and treatment condition. Rows illustrate data for (a and b) LH pulse mass, (c and d) mean LH, and (e and f) mean FSH. (a, c, and e) (left) Individual mean data (denoted by connected ○) with placebo (PBO) and progesterone (P4) during wake blocks; (right) individual mean data during sleep blocks. (b, d, and f) (left) Mean data (denoted by ●) for PBO and P4 conditions during wake blocks; (right) mean data during sleep blocks. Vertical lines identify 95% confidence intervals for the mean. P values relate to changes between PBO and P4 conditions. Although not significant, P values relating to the interaction between treatment condition and sleep status are shown. Figure 4. View largeDownload slide Secondary outcomes: LH pulse mass, mean LH, and mean FSH as a function of sleep status and treatment condition. Rows illustrate data for (a and b) LH pulse mass, (c and d) mean LH, and (e and f) mean FSH. (a, c, and e) (left) Individual mean data (denoted by connected ○) with placebo (PBO) and progesterone (P4) during wake blocks; (right) individual mean data during sleep blocks. (b, d, and f) (left) Mean data (denoted by ●) for PBO and P4 conditions during wake blocks; (right) mean data during sleep blocks. Vertical lines identify 95% confidence intervals for the mean. P values relate to changes between PBO and P4 conditions. Although not significant, P values relating to the interaction between treatment condition and sleep status are shown. Although mean LH was approximately twofold higher with progesterone during waking and sleep hours, these differences were not statistically significant [Fig. 4(c) and 4(d)]. Mean FSH was similar under progesterone and placebo conditions during wake periods and sleep periods [Fig. 4(e) and 4(f)]. Mean LH and mean FSH were not predicted by sleep status, sequence order, or crossover period, and there was no interaction between treatment condition and sleep status for either. Post hoc analyses Compared with the placebo condition, progesterone was associated with significantly lower LH pulse frequency during both evening wake (1900 to 2300 hours) and morning wake (0700 to 1100 hours) time blocks, with no differences observed during either sleep-related time block [(Fig. 5(a)]. The Spearman’s rank correlation procedure did not disclose a significant correlation between (1) percent change in waking LH pulse frequency related to progesterone administration and (2) baseline free testosterone concentration in this cohort of late pubertal girls [rs = 0.291, P = 0.386; Fig. 5(b)]. Also of note, an appraisal of the data partitioned by obesity status suggested that study results were similar in obese and nonobese subjects (see Supplemental Materials). Figure 5. View largeDownload slide Post hoc analyses. (a) Mean LH pulse frequency with placebo (PBO) and progesterone (P4) in each 4-hour time block. Wake time blocks are 1900 to 2300 and 0700 to 1100 hours; sleep time blocks are 2300 to 0300 and 0300 to 0700 hours. Box-and-whisker plots denote the median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). (b) The relationship between the percent change in waking LH pulse frequency related to P4 administration and free testosterone measured during the PBO admission (calculated using SHBG and total testosterone measured by LC-MS/MS). Figure 5. View largeDownload slide Post hoc analyses. (a) Mean LH pulse frequency with placebo (PBO) and progesterone (P4) in each 4-hour time block. Wake time blocks are 1900 to 2300 and 0700 to 1100 hours; sleep time blocks are 2300 to 0300 and 0300 to 0700 hours. Box-and-whisker plots denote the median (line inside the box); mean (open square); 25th and 75th percentiles (bottom and top of box); and minimum and maximum (bottom and top whiskers). (b) The relationship between the percent change in waking LH pulse frequency related to P4 administration and free testosterone measured during the PBO admission (calculated using SHBG and total testosterone measured by LC-MS/MS). Discussion The current study provides compelling evidence that in late pubertal girls without hyperandrogenism, progesterone acutely suppresses waking LH pulse frequency to a greater extent than sleep-associated LH pulse frequency. These data support our earlier findings in early-to-midpubertal girls (9), and taken together, these two studies strongly suggest that the GnRH pulse generator in pubertal girls is differentially sensitive to progesterone inhibition based on sleep-wake status. We hypothesize that this phenomenon contributes to the evolution of sleep-wake LH pulse patterns across puberty in girls. As previously described, waking LH pulse frequency gradually increases across puberty, whereas sleep-associated LH frequency remains relatively constant: in early puberty, LH pulse frequency increases to ∼0.5 pulses/h during sleep, but by late puberty, LH pulse frequency decreases to ∼0.5 pulses/h during sleep (3, 10). Mechanisms underlying the evolution of such patterns are unknown. However, progesterone is a primary modulator of day-to-day GnRH pulse frequency slowing in women (17, 18), and contrary to girls with spontaneous puberty, age-matched girls with gonadal dysgenesis demonstrate no sleep-wake differences in LH pulse frequency (19), implying a role for sex steroids. Our previous study in early pubertal girls (9) suggested that exogenous progesterone quickly suppresses LH pulse frequency while awake but not during sleep. We thus hypothesized that waking GnRH frequency is regulated by a complement of neuronal inputs that are sensitive to progesterone negative feedback but that sleep-related GnRH pulse frequency is regulated by an ensemble of neuronal inputs that, as a unit, is relatively stereotyped and not readily influenced by low levels of progesterone (1). The results of our current study strongly support this hypothesis. We note that other studies have suggested differential control of sleep vs wake LH pulse frequency: in normal adult women studied in the late follicular phase, dietary calorie restriction preferentially reduced daytime LH pulse frequency (20, 21). According to our overall working model (1)—represented in schematic form in Supplemental Materials—waking and sleep-associated GnRH pulse frequency are regulated by different complements of neuronal inputs: the former is responsive to progesterone negative feedback, but the latter is relatively resistant to such inhibition. Very low waking GnRH pulse frequency in early puberty relates to exquisite sensitivity to progesterone negative feedback. [We suggest that in normal early pubertal girls (i.e., before the establishment of cyclic ovulation), high nocturnal frequency may be necessary to enhance LH secretion, which promotes ovarian sex steroid production, whereas a low daytime pulse frequency may be critical to maintain FSH secretion, which supports follicular development.] Androgen concentrations increase across puberty (10, 22, 23), and androgens antagonize progesterone negative feedback (24, 25): our working model holds that the gradual rise of testosterone across puberty reduces the ability of progesterone to inhibit waking GnRH pulse frequency, thus allowing a gradual increase in waking pulse frequency. However, outside of the luteal phase, sleep-associated GnRH pulse frequency remains relatively constant across puberty, as such inputs are not readily altered by low progesterone levels. We have also previously proposed that our working model may help explain the genesis of neuroendocrine abnormalities in girls with hyperandrogenemia and nascent PCOS (1): if pubertal initiation (GnRH pulse generator reawakening) occurs in the setting of relative hyperandrogenemia, then the higher androgen concentrations would antagonize progesterone inhibition of waking GnRH pulse frequency, leading to a rapid transition to high GnRH pulse frequency during both day and night. This pattern would exaggerate LH release and further promote androgen production while also limiting FSH synthesis and, thus, follicular development. We are currently pursuing studies to test these hypotheses. Whereas the post hoc analysis in the current study did not suggest a significant correlation between progesterone-induced changes in waking LH pulse frequency and free testosterone concentrations, we suggest that this may reflect relatively low numbers of subjects who were all within a narrow developmental range: inclusion of a greater number of subjects exhibiting a broader range of testosterone values may be required to demonstrate this putative relationship. Of interest, total testosterone by LC-MS/MS was 20% lower with progesterone administration compared with the placebo condition. Although progesterone reduced LH pulse frequency while awake, LH pulse mass simultaneously increased, and mean LH concentrations did not change. This may suggest that waking LH pulse frequency per se is an important determinant of ovarian androgen production, and it is consistent with the notion that an inability to suppress waking LH pulse frequency appropriately in early puberty may promote hyperandrogenemia and support a progression to PCOS. Our prior study in early pubertal girls suggested that progesterone abolishes waking LH pulse secretion within 3 to 7 hours (9), and our current study suggests that progesterone reduces waking LH pulse frequency by 26% within 12 to 16 hours in later puberty. In contrast, our previous studies in adult women suggest that progesterone suppresses waking LH pulse frequency more slowly (14, 26). We hypothesize that these apparently discordant observations reflect differences in androgen exposure: the effects of progesterone negative feedback are rapid and profound when androgen concentrations are very low, but they are less profound—or perhaps require longer periods of time—when androgen concentrations are higher. We also note that progesterone will suppress sleep-related LH pulse frequency when progesterone exposure is sufficiently high and long, as occurs in the luteal phase (27, 28). Progesterone administration given for 7 to 10 days, with or without concomitant estradiol administration, will also decrease sleep-related LH pulse frequency in adult women (26, 29, 30) and adolescent girls (31, 32), and 7 days of progesterone administration slows sleep-related, free alpha-subunit (GnRH) pulse frequency in estradiol-replaced postmenopausal women (33). It remains unknown whether progesterone has a differential effect on wake vs sleep LH pulse frequency in adult women, either in terms of degree of suppression or rapidity of suppression. Progesterone action at the hypothalamus appears to require the permissive presence of estradiol (34, 35), presumably because estradiol can increase progesterone receptor expression in the hypothalamus (36, 37). Although the minimum estradiol level needed for effective progesterone action is not known, progesterone acutely reduced LH pulse frequency during waking hours in the current study, suggesting that estradiol concentrations of 27.7 ± 19.9 pg/mL (mean ± SD) were sufficient to induce and/or maintain the relevant hypothalamic progesterone receptors in late pubertal girls. It remains possible that effective progesterone action requires higher estradiol concentrations in other groups (e.g., normal adult women, women with PCOS). Our data showed that LH pulse mass increased with progesterone administration regardless of sleep status. This is in keeping with earlier studies: progesterone increases gonadotropin responses to exogenous GnRH (38–40) and augments gonadotropin secretion in estradiol-pretreated women (14, 16, 41). Although mean LH was much higher with progesterone compared with placebo in three subjects, progesterone did not have a statistically significant overall effect on mean LH or mean FSH. The current study addresses several limitations of our prior study in early pubertal girls (9). First, a relatively small number of subjects (n = 5) received progesterone in our earlier study, and they were compared with a separate group who did not receive progesterone (n = 13). The current study involved 11 subjects evaluated under both progesterone and placebo conditions; the crossover design allowed each girl to serve as her own control, avoiding potential confounding related to between-group differences. Second, it seemed possible (albeit unlikely) that all girls receiving progesterone in our earlier study would have had undetectable wake-associated LH pulse frequency regardless of progesterone administration. In contrast to early pubertal girls, postmenarcheal girls are reliably expected to exhibit daytime LH pulses, and their higher LH pulse amplitudes enhance the reliability of LH pulse detection. Third, blood withdrawal limits in early pubertal girls constrained prior surveillance of daytime LH secretion to 4 hours; in the current study, greater blood volumes permitted frequent sampling during an 8-hour sleep period and for a full 10 hours while awake. Lastly, sleep was not formally evaluated in our earlier study, but we used actigraphy to determine sleep periods in the current study. Regarding potential limitations of this study, we recognize that the use of time blocks as surrogates for wake and sleep periods can be imprecise; however, LH pulse frequency for actigraphy-directed sleep periods correlated very highly with time block-derived estimates for sleep-related LH pulse frequency (rs = 0.977), and the results of actigraphy-directed analysis were virtually identical to the results of time block-directed analysis. Second, we assessed waking LH pulse frequency as the average of two, 4-hour time blocks that were temporally separated, instead of assessment during a continuous 8-hour window (as for sleep-related LH pulse frequency). This decision reflected a desire to minimize the potential impact of time-related confounding: if we had assessed waking pulse frequency from 1500 to 2300 hours or from 0700 to 1500 hours, then sleep-associated pulse frequency would have been associated with progesterone exposure that was 8 hours longer or shorter, respectively. Third, progesterone concentrations achieved in this study were supraphysiologic compared with the normal follicular phase, but this was necessary. Recapitulation of physiological progesterone concentrations would not be expected to alter LH secretion vis-à-vis the placebo condition. Fourth, we studied late pubertal girls to investigate a physiological phenomenon that may be most relevant in early pubertal subjects. However, early pubertal subjects are more difficult to recruit; they less reliably exhibit LH secretion while awake; and their lower body weight translates to more restrictive blood withdrawal limitations, which constrain the duration of LH surveillance and discourage a two-admission crossover design. Lastly, although most subjects had regular menses, we did not directly assess ovulatory status for the cycles immediately preceding overnight admissions, and it remains possible that differences in ovulatory status (e.g., ovulatory vs anovulatory) preceding overnight admissions could have influenced study results. Given the study’s crossover design, where each subject serves as her own control, we assume that any potential confounding would primarily pertain to unequal within-subject study conditions (e.g., if a subject’s first study admission was preceded by an ovulatory cycle, and her second study admission was preceded by an anovulatory cycle). However, randomization of treatment order would be expected to mitigate potential bias in this regard: any bias would occur in one direction in roughly one-half of the subjects (e.g., exaggerating or masking the apparent effect of progesterone) and in the opposite direction in roughly one-half of the subjects. Thus, whereas this putative confounder could have contributed to study outcome variability, it seems unlikely that it could have systematically biased our results or led to type 1 statistical error. In conclusion, this highly rigorous clinical research study demonstrates that progesterone acutely decreases waking LH pulse frequency to a greater extent than sleep-associated LH frequency in late pubertal girls without hyperandrogenemia. Taken together with our prior study in early-to-midpubertal girls, these data strongly support the hypothesis that the pubertal GnRH pulse generator is differentially sensitive to progesterone negative feedback depending on sleep status. This may represent an important mechanism directing day-night changes of GnRH pulse frequency during normal puberty. Abbreviations: CRU Clinical Research Unit CV coefficient of variation FSH follicle-stimulating hormone GnRH gonadotropin-releasing hormone LC-MS/MS liquid chromatography-tandem mass spectrometry LH luteinizing hormone LM linear-mixed PCOS polycystic ovary syndrome RIA radioimmunoassay rs Spearman’s rank correlation coefficient SD standard deviation SHBG sex hormone-binding globulin UVA University of Virginia. Acknowledgments We gratefully acknowledge Anne Gabel, Katherine Ehrlich, Deborah M. Sanderson, Melissa Gilrain, Amy Anderson, and Eleanor Hutchens for subject recruitment, study scheduling, and assistance with data management. We also extend our gratitude to the nurses and staff of the Clinical Research Unit at University of Virginia for implementation of the sampling protocols and to Dan Haisenleder and the Center for Research in Reproduction Ligand Core Laboratory for performance of assays. Financial Support: This work was supported by US National Institutes of Health (NIH) Grant R01 HD058671 (to C.R.M. and J.T.P.); Eunice Kennedy Shriver National Institute of Child Health and Human Development/NIH through Cooperative Agreement P50 HD28934 as part of the National Centers for Translational Research in Reproduction and Infertility (to C.R.M., J.C.M., C.M.B.S., and J.T.P.); and NIH Grants F32 HD088047 (to S.H.K.), F32 HD091951 (to J.A.L.), F32 HD078088 (to R.B.), F32 HD066855 (to J.S.C.), and K23 HD070854 (to C.M.B.S.). Clinical Trial Information: ClinicalTrials.gov no. NCT00929006 (registered 26 June 2009). Author Contributions: S.H.K. was the primary individual performing data collection, organization, and analysis, and she was the primary author of manuscript. J.A.L., R.B., and J.S.C. contributed to study design and implementation, including subject recruitment, oversight of CRU admissions, and data collection/analysis, and they provided editorial guidance during manuscript preparation. J.T.P. established all statistical plans before study initiation, provided the randomization schedule to the Investigational Pharmacy, and performed all primary and secondary statistical analyses described herein, and he provided editorial guidance during manuscript preparation. C.M.B.S. and J.C.M. provided input into study design and data interpretation, and they provided editorial guidance during manuscript preparation. C.R.M. was the principal investigator of the study and was responsible for study design and oversight. He supervised data collection and organization, performed all pulse detection analyses, directed all data analysis, and materially participated in manuscript preparation. Disclosure Summary: The authors have nothing to disclose. References 1. McCartney CR. Maturation of sleep-wake gonadotrophin-releasing hormone secretion across puberty in girls: potential mechanisms and relevance to the pathogenesis of polycystic ovary syndrome. J Neuroendocrinol . 2010; 22( 7): 701– 709. Google Scholar PubMed  2. Boyar R, Finkelstein J, Roffwarg H, Kapen S, Weitzman E, Hellman L. Synchronization of augmented luteinizing hormone secretion with sleep during puberty. N Engl J Med . 1972; 287( 12): 582– 586. Google Scholar CrossRef Search ADS PubMed  3. Apter D, Bützow TL, Laughlin GA, Yen SS. Gonadotropin-releasing hormone pulse generator activity during pubertal transition in girls: pulsatile and diurnal patterns of circulating gonadotropins. J Clin Endocrinol Metab . 1993; 76( 4): 940– 949. Google Scholar PubMed  4. Shaw ND, Butler JP, McKinney SM, Nelson SA, Ellenbogen JM, Hall JE. Insights into puberty: the relationship between sleep stages and pulsatile LH secretion. J Clin Endocrinol Metab . 2012; 97( 11): E2055– E2062. Google Scholar CrossRef Search ADS PubMed  5. Burger LL, Haisenleder DJ, Dalkin AC, Marshall JC. Regulation of gonadotropin subunit gene transcription. J Mol Endocrinol . 2004; 33( 3): 559– 584. Google Scholar CrossRef Search ADS PubMed  6. Thompson IR, Kaiser UB. GnRH pulse frequency-dependent differential regulation of LH and FSH gene expression. Mol Cell Endocrinol . 2014; 385( 1-2): 28– 35. Google Scholar CrossRef Search ADS PubMed  7. Burt Solorzano CM, Beller JP, Abshire MY, Collins JS, McCartney CR, Marshall JC. Neuroendocrine dysfunction in polycystic ovary syndrome. Steroids . 2012; 77( 4): 332– 337. Google Scholar CrossRef Search ADS PubMed  8. McCartney CR, Blank SK, Prendergast KA, Chhabra S, Eagleson CA, Helm KD, Yoo R, Chang RJ, Foster CM, Caprio S, Marshall JC. Obesity and sex steroid changes across puberty: evidence for marked hyperandrogenemia in pre- and early pubertal obese girls. J Clin Endocrinol Metab . 2007; 92( 2): 430– 436. Google Scholar CrossRef Search ADS PubMed  9. Collins JS, Marshall JC, McCartney CR. Differential sleep-wake sensitivity of gonadotropin-releasing hormone secretion to progesterone inhibition in early pubertal girls. Neuroendocrinology . 2012; 96( 3): 222– 227. Google Scholar CrossRef Search ADS PubMed  10. McCartney CR, Prendergast KA, Blank SK, Helm KD, Chhabra S, Marshall JC. Maturation of luteinizing hormone (gonadotropin-releasing hormone) secretion across puberty: evidence for altered regulation in obese peripubertal girls. J Clin Endocrinol Metab . 2009; 94( 1): 56– 66. Google Scholar CrossRef Search ADS PubMed  11. Wierman ME, Auchus RJ, Haisenleder DJ, Hall JE, Handelsman D, Hankinson S, Rosner W, Singh RJ, Sluss PM, Stanczyk FZ. Editorial: The new instructions to authors for the reporting of steroid hormone measurements. J Clin Endocrinol Metab . 2014; 99( 12): 4375. Google Scholar CrossRef Search ADS PubMed  12. Bhasin S, Pencina M, Jasuja GK, Travison TG, Coviello A, Orwoll E, Wang PY, Nielson C, Wu F, Tajar A, Labrie F, Vesper H, Zhang A, Ulloor J, Singh R, D’Agostino R, Vasan RS. Reference ranges for testosterone in men generated using liquid chromatography tandem mass spectrometry in a community-based sample of healthy nonobese young men in the Framingham Heart Study and applied to three geographically distinct cohorts. J Clin Endocrinol Metab . 2011; 96( 8): 2430– 2439. Google Scholar CrossRef Search ADS PubMed  13. Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab . 1999; 84( 10): 3666– 3672. Google Scholar CrossRef Search ADS PubMed  14. Hutchens EG, Ramsey KA, Howard LC, Abshire MY, Patrie JT, McCartney CR. Progesterone has rapid positive feedback actions on LH release but fails to reduce LH pulse frequency within 12 h in estradiol-pretreated women. Physiol Rep . 2016; 4( 16): e12891. Google Scholar CrossRef Search ADS PubMed  15. Johnson ML, Pipes L, Veldhuis PP, Farhy LS, Boyd DG, Evans WS. AutoDecon, a deconvolution algorithm for identification and characterization of luteinizing hormone secretory bursts: description and validation using synthetic data. Anal Biochem . 2008; 381( 1): 8– 17. Google Scholar CrossRef Search ADS PubMed  16. McCartney CR, Blank SK, Marshall JC. Progesterone acutely increases LH pulse amplitude but does not acutely influence nocturnal LH pulse frequency slowing during the late follicular phase in women. Am J Physiol Endocrinol Metab . 2007; 292( 3): E900– E906. Google Scholar CrossRef Search ADS PubMed  17. McCartney CR, Marshall JC. Neuroendocrinology of reproduction. In: Strauss JF, Barbieri RL, eds. Yen & Jaffe's Reproductive Endocrinology: Physiology, Pathophysiology, and Clinical Management . 7th ed. Philadelphia, PA: Elsevier Saunders, 2014: 3– 26 18. Hall JE. Neuroendocrine control of the menstrual cycle. In: Strauss JF, Barbieri RL, eds. Yen & Jaffe's Reproductive Endocrinology: Physiology, Pathophysiology, and Clinical Management . 7th ed. Philadelphia, PA: Elsevier Saunders, 2014: 141– 156 19. Cemeroglu AP, Foster CM, Warner R, Kletter GB, Marshall JC, Kelch RP. Comparison of the neuroendocrine control of pubertal maturation in girls and boys with spontaneous puberty and in hypogonadal girls. J Clin Endocrinol Metab . 1996; 81( 12): 4352– 4357. Google Scholar PubMed  20. Loucks AB, Heath EM. Dietary restriction reduces luteinizing hormone (LH) pulse frequency during waking hours and increases LH pulse amplitude during sleep in young menstruating women. J Clin Endocrinol Metab . 1994; 78( 4): 910– 915. Google Scholar PubMed  21. Loucks AB, Verdun M, Heath EM. Low energy availability, not stress of exercise, alters LH pulsatility in exercising women. J Appl Physiol (1985) . 1998; 84( 1): 37– 46. Google Scholar CrossRef Search ADS PubMed  22. Ankarberg C, Norjavaara E. Diurnal rhythm of testosterone secretion before and throughout puberty in healthy girls: correlation with 17beta-estradiol and dehydroepiandrosterone sulfate. J Clin Endocrinol Metab . 1999; 84( 3): 975– 984. Google Scholar PubMed  23. Mitamura R, Yano K, Suzuki N, Ito Y, Makita Y, Okuno A. Diurnal rhythms of luteinizing hormone, follicle-stimulating hormone, testosterone, and estradiol secretion before the onset of female puberty in short children. J Clin Endocrinol Metab . 2000; 85( 3): 1074– 1080. Google Scholar PubMed  24. Eagleson CA, Gingrich MB, Pastor CL, Arora TK, Burt CM, Evans WS, Marshall JC. Polycystic ovarian syndrome: evidence that flutamide restores sensitivity of the gonadotropin-releasing hormone pulse generator to inhibition by estradiol and progesterone. J Clin Endocrinol Metab . 2000; 85( 11): 4047– 4052. Google Scholar PubMed  25. Pielecka J, Quaynor SD, Moenter SM. Androgens increase gonadotropin-releasing hormone neuron firing activity in females and interfere with progesterone negative feedback. Endocrinology . 2006; 147( 3): 1474– 1479. Google Scholar CrossRef Search ADS PubMed  26. Pastor CL, Griffin-Korf ML, Aloi JA, Evans WS, Marshall JC. Polycystic ovary syndrome: evidence for reduced sensitivity of the gonadotropin-releasing hormone pulse generator to inhibition by estradiol and progesterone. J Clin Endocrinol Metab . 1998; 83( 2): 582– 590. Google Scholar PubMed  27. Filicori M, Santoro N, Merriam GR, Crowley WF Jr. Characterization of the physiological pattern of episodic gonadotropin secretion throughout the human menstrual cycle. J Clin Endocrinol Metab . 1986; 62( 6): 1136– 1144. Google Scholar CrossRef Search ADS PubMed  28. Rossmanith WG, Liu CH, Laughlin GA, Mortola JF, Suh BY, Yen SS. Relative changes in LH pulsatility during the menstrual cycle: using data from hypogonadal women as a reference point. Clin Endocrinol (Oxf) . 1990; 32( 5): 647– 660. Google Scholar CrossRef Search ADS PubMed  29. Soules MR, Steiner RA, Clifton DK, Cohen NL, Aksel S, Bremner WJ. Progesterone modulation of pulsatile luteinizing hormone secretion in normal women. J Clin Endocrinol Metab . 1984; 58( 2): 378– 383. Google Scholar CrossRef Search ADS PubMed  30. McCartney CR, Gingrich MB, Hu Y, Evans WS, Marshall JC. Hypothalamic regulation of cyclic ovulation: evidence that the increase in gonadotropin-releasing hormone pulse frequency during the follicular phase reflects the gradual loss of the restraining effects of progesterone. J Clin Endocrinol Metab . 2002; 87( 5): 2194– 2200. Google Scholar CrossRef Search ADS PubMed  31. Chhabra S, McCartney CR, Yoo RY, Eagleson CA, Chang RJ, Marshall JC. Progesterone inhibition of the hypothalamic gonadotropin-releasing hormone pulse generator: evidence for varied effects in hyperandrogenemic adolescent girls. J Clin Endocrinol Metab . 2005; 90( 5): 2810– 2815. Google Scholar CrossRef Search ADS PubMed  32. Blank SK, McCartney CR, Chhabra S, Helm KD, Eagleson CA, Chang RJ, Marshall JC. Modulation of gonadotropin-releasing hormone pulse generator sensitivity to progesterone inhibition in hyperandrogenic adolescent girls--implications for regulation of pubertal maturation. J Clin Endocrinol Metab . 2009; 94( 7): 2360– 2366. Google Scholar CrossRef Search ADS PubMed  33. Gill S, Lavoie HB, Bo-Abbas Y, Hall JE. Negative feedback effects of gonadal steroids are preserved with aging in postmenopausal women. J Clin Endocrinol Metab . 2002; 87( 5): 2297– 2302. Google Scholar CrossRef Search ADS PubMed  34. Karsch FJ, Weick RF, Hotchkiss J, Dierschke DJ, Knobil E. An analysis of the negative feedback control of gonadotropin secretion utilizing chronic implantation of ovarian steroids in ovariectomized rhesus monkeys. Endocrinology . 1973; 93( 2): 478– 486. Google Scholar CrossRef Search ADS PubMed  35. Nippoldt TB, Reame NE, Kelch RP, Marshall JC. The roles of estradiol and progesterone in decreasing luteinizing hormone pulse frequency in the luteal phase of the menstrual cycle. J Clin Endocrinol Metab . 1989; 69( 1): 67– 76. Google Scholar CrossRef Search ADS PubMed  36. MacLusky NJ, McEwen BS. Oestrogen modulates progestin receptor concentrations in some rat brain regions but not in others. Nature . 1978; 274( 5668): 276– 278. Google Scholar CrossRef Search ADS PubMed  37. Romano GJ, Krust A, Pfaff DW. Expression and estrogen regulation of progesterone receptor mRNA in neurons of the mediobasal hypothalamus: an in situ hybridization study. Mol Endocrinol . 1989; 3( 8): 1295– 1300. Google Scholar CrossRef Search ADS PubMed  38. Chang RJ, Jaffe RB. Progesterone effects on gonadotropin release in women pretreated with estradiol. J Clin Endocrinol Metab . 1978; 47( 1): 119– 125. Google Scholar CrossRef Search ADS PubMed  39. Nippoldt TB, Khoury S, Barkan A, Kelch RP, Marshall JC. Gonadotrophin responses to GnRH pulses in hypogonadotrophic hypogonadism: LH responsiveness is maintained in the presence of luteal phase concentrations of oestrogen and progesterone. Clin Endocrinol (Oxf) . 1987; 26( 3): 293– 301. Google Scholar CrossRef Search ADS PubMed  40. Couzinet B, Brailly S, Bouchard P, Schaison G. Progesterone stimulates luteinizing hormone secretion by acting directly on the pituitary. J Clin Endocrinol Metab . 1992; 74( 2): 374– 378. Google Scholar PubMed  41. Liu JH, Yen SS. Induction of midcycle gonadotropin surge by ovarian steroids in women: a critical evaluation. J Clin Endocrinol Metab . 1983; 57( 4): 797– 802. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off