Gender, Sex-Steroid, and Secretagogue-Selective Recovery from Growth Hormone-Induced Feedback in Older Women and Men

Abstract

Context:

GH negatively regulates its own secretion. How gender, sex steroids, and secretagogues modulate GH autofeedback is not known.

Hypothesis/Objective:

Supplementation with sex steroids and/or a peptidyl secretagogue will enhance the escape of GH from autoinhibition, thus framing a mechanism for amplifying pulsatile GH secretion.

Subjects and Setting:

Ten healthy postmenopausal women and 10 comparably aged men participated at the Clinical-Translational Science Unit.

Design/Interventions:

Randomly ordered, double-blind, prospective crossover treatment with placebo vs. testosterone (men) or placebo vs. estradiol (women). Autofeedback was imposed by an iv pulse of GH. Recovery of feedback inhibition was quantified during constant infusion of saline, GHRH, or GH-releasing peptide-2 (three peptide categories).

Outcomes/Results:

During negative feedback, total (integrated) GH recovery depended upon gender (P = 0.017), sex hormone (P < 0.001), and peptide category (P < 0.001). Mechanistic analysis revealed that feedback-suppressed nadir GH concentrations were determined by sex-steroid treatment (P = 0.018) but not by gender (P = 0.444). Peak GH escape was controlled by both treatment (P = 0.004) and gender (P = 0.003). Nadir GH and peak GH during feedback were enhanced by GHRH or GHRP-2 (P < 0.001 for both). Gender × peptide (P = 0.012 for nadir GH), treatment × peptide (P < 0.001 total and peak GH), and gender × treatment (P = 0.017 nadir GH) regulated GH recovery interactively.

Conclusion:

Gender, sex-steroid supplementation, and secretagogue type confer distinct feedback-rescuing effects, introducing a new level of complexity in the control of pulsatile GH regulation.

Protein hormones, such as GH, are secreted by the anterior pituitary gland primarily (>85%) in discrete bursts (1, 2). Laboratory experiments and mathematical modeling of GH pulse-renewal mechanisms highlight the importance of reversible, time-delayed feed-forward (stimulatory) and feedback (inhibitory) interactions (3). How such dynamic mechanisms are regulated in the human is unknown. Among relevant effectors, the 44-amino-acid hypothalamic peptide GHRH is a primary proximate agonist of GH secretion in vitro and in vivo (4). In contradistinction, ghrelin is an endogenous 28-amino-acid Ser³-octanoylated GH-releasing peptide (GHRP) of gastro-pancreatico-hypothalamo-pituitary origin (5, 6). The acylated peptide and synthetic analogs, such as GHRP-2, stimulate GH secretion by 5- to 20-fold alone and by 30- to 120-fold in synergy with GHRH (6, 7). The two peptidyl signals are important in physiological control, because inactivating mutations of the pituitary GHRH receptor and transgenic silencing of the brain GHRP receptor can reduce GH and IGF-I concentrations significantly (6, 8).

In healthy adults, bolus infusion of GHRH or ghrelin (GHRP) stimulates burst-like release of GH (6–12). Stimulation is antagonized by the hypothalamic tetradecapeptide, somatostatin (SS), which also mediates negative feedback by GH and IGF-I (13). Hypothalamic SS outflow inhibits pituitary exocytosis of GH and represses hypothalamic GHRH release (14). The conjoint pituitary and hypothalamic effects of GH-induced SS outflow suppress pulsatile GH secretion in a reversible time-delimited fashion (15). A model-based prediction, albeit untested to date, is that sex-steroid supplementation could augment pulsatile GH secretion by attenuating GH-induced SS-mediated feedback restraint (2). If this hypothesis is true, then failure of such disinhibition might contribute to reduced pulsatile GH secretion in a low sex-steroid milieu or in aging individuals (9, 16).

The present investigation tests the hypothesis that supplementation with estradiol (E₂) in women and testosterone (T) in men compared with placebo will recover pulsatile GH secretion from GH feedback-enforced repression and that recovery will be determined by gender and peptidyl secretagogue type.

Subjects and Methods

Human subjects

Participants provided Mayo Institutional Review Board-approved written, voluntary informed consent and a detailed medical history and underwent a screening physical examination as outpatients. Biochemical screening was performed to ensure normal hepatic, renal, hematological, metabolic, and endocrine function before admission to the study.

Criteria for involvement or exclusion

Included in the study were community-dwelling, ambulatory, and healthy 50- to 80-yr-old women (n = 10) and men (n = 10), who provided voluntary written informed Institutional Review Board-approved consent.

Exclusion criteria included premenopausal status (screening FSH concentration < 30 IU/liter and E₂ > 35 pg/ml); concurrent use of neuroactive medications; sex hormones; systemic illness; diabetes mellitus; untreated endocrinopathy; abnormal medical history, examination, or biochemical screening data; acute or chronic organ (including inflammatory) disease; drug or alcohol abuse; hemoglobin below 11.5 g/dl in women and below 12 g/dl in men; history of thrombotic arterial disease (stroke, transient ischemic attack, myocardial infarction, or angina) or deep-venous thrombophlebitis; history or suspicion of breast cancer; prostatic disease (elevated prostate-specific antigen, indeterminate nodule or mass, carcinoma, obstructive uropathy); anticoagulant use in men (due to im T injections); allergy to peanut oil in men (peanut oil may be an excipient for T); and unwillingness to provide written informed consent.

Recruitment

Reimbursement was for time spent in the study.

Study schedule

Each subject received randomly ordered placebo vs. sex-steroid hormone supplementation for 21 d at least 6 wk apart. GH autofeedback and peptidyl stimuli were randomized in a within-subject crossover design with three separate 8-h infusion sessions in each sex-steroid milieu (total of six sessions per subject).

In women, an oral E₂-repletion schedule was used to mimic concentrations attained in the young-adult late follicular phase, viz., 1.0 mg oral micronized E₂ twice daily for 21 d. Medroxyprogesterone 5 mg was given orally for 12 d after the study, in accordance with good standards of clinical practice.

In men, T enanthate 200 mg was injected im on d 1, 8, and 15 in the outpatient setting. This schedule of androgen supplementation doubles mean (24-h) GH concentrations and elevates IGF-I concentrations by 35% in older, but not young, men (17).

Sampling and infusion

Participants were admitted to the inpatient Clinical-Translational Research Unit by 1730 h the evening before study. To limit nutritional confounds, a vegetarian or nonvegetarian standardized meal (8 kcal/kg for women and 10 kcal/kg for men of 50% carbohydrate, 20% protein, and 30% fat) was served at 1800 h. Subjects then remained fasting, alcohol-abstinent, and caffeine-free until 1400 h the next day.

GH secretion was monitored by sampling blood (1 ml) every 10 min for 8 h in the fasting state (0800–1600 h) beginning 30 min before iv GH injection (0830 h) and continuing during randomly ordered double-blind infusions of saline or a peptide secretagogue (Fig. 1). Sessions were scheduled during the time window including d 10–21 (where d 1 is the onset of placebo vs. sex-hormone administration). Peptide infusions were as follows: 1) continuous iv infusion of saline (20 ml/h) from 0800–1600 h with a single superimposed 6-min iv infusion of GH (1 μg/kg) delivered at 0830–0836 h, 2) constant iv infusion of GHRH (1 μg/kg · h) during the inclusive time window 0800–1600 h with the same superimposed GH pulse at 0830 h, and 3) continuous iv infusion of GHRP-2 (1.0 μg/kg · h) from 0800–1600 h with the same superimposed GH pulse at 0830 h.

Fig. 1.

Schema of experimental design comprising bolus iv GH (1 μg/kg) injection at 0830 h, repetitive 10-min sampling for 8 h between 0800 and 1600 h, and superimposed continuous iv infusion of saline (20 ml/h), GHRH (1 μg/kg · h), or GHRP-2 (1 μg/kg · h). Each subject was studied without (placebo) and with (T or E₂) sex-steroid supplementation (six sessions in total).

Subjects undertook each of the three secretagogue paradigms during both placebo and sex-steroid supplementation (yielding six sessions per person and 120 study sessions altogether).

Analytical methods

Serum concentrations of GH were determined via sensitive, precise, specific and previously validated two-site monoclonal immunochemiluminescence assay (sensitivity 0.010 μg/liter), as reported (18, 19). Rare GH values below the assay threshold (0.01 μg/liter) were assigned a probabilistic value of 0.005 μg/liter rather than zero. All other assays have been described recently (20, 21).

Biostatistical analysis

The principal a priori postulate was that gender, sex steroids, and peptidyl drive augment post-feedback escape of integrated and more specifically nadir and peak GH concentrations in healthy older adults. Integrated GH during feedback escape was the trapezoidal area under the 5.5-h GH-recovery curve beginning 120 min after bolus iv GH injection (viz., 1030–1600 h). The nadir was defined as the lowest and the peak the highest three-point moving-average GH concentration over the 6-h interval beginning 90 min (5.6 half-lives) after the iv pulse of GH. This was because we desired an accurate GH-suppressed nadir and a GH-recovered peak, unaffected by the decay of injected GH. Effects of the three main (categorical) factors gender (two factors), sex-steroid (two factors), and peptide (three factors) were assessed by three-way mixed-effects analysis of covariance (ANCOVA) of natural-logarithmically transformed values. Log transformation was employed to produce symmetric measurement distributions and comparable measurement variability, given the statistical assumption of similar residual variation within all treatment groups. The saline/placebo outcome was the covariate. ANCOVA model parameters were estimated by residual maximum likelihood assuming compound symmetry of the variance-covariance matrix. Post hoc comparisons were made by Tukey's honestly significantly different (HSD) test at experiment-wise two-tailed (α) P < 0.05. Double- and triple-factor interactions were assessed concomitantly. Anticipated mean (±sd) effect sizes over saline for peptide infusions were as for GHRP-2, 5.2 ± 0.48 (n = 18), and GHRH, 4.9 ± 0.53 (n = 28), and over placebo for T or E₂ administration, 2.1 ± 0.18-fold a9n = 43). Based upon the foregoing estimates, approximate statistical power exceeded 90% to detect 40% amplification by T or E₂ of the GH nadir or peak at P < 0.05, if 10 men and 10 women each completed all six admissions (22). SYSTAT-11 software (Richmond, CA) was applied in all analyses.

Data are presented as geometric mean ± sem, unless specified otherwise.

Results

All 20 volunteers completed all six infusion/sampling sessions. Baseline data in the 10 men and 10 women are given in Supplemental Table 1 (published on The Endocrine Society's Journals Online web site at http://jcem.endojournals.org). Body mass index and age were comparable in the sexes. By unpaired Student's t test, sex affected E₂ (lower in postmenopausal women than older men), T (higher in men), and SHBG, LH, and FSH (higher in women), but not IGF-I and prolactin concentrations.

Eight-hour GH-concentration profiles (algebraic mean ± sem) obtained by sampling blood every 10 min fasting in the 10 men and 10 women are shown in Fig. 2. There was marked feedback inhibition after the iv GH pulse in the placebo/saline setting and visually evident enhanced recovery variously by E₂/T, GHRH, and GHRP-2. Mean baseline (preinjected) GH concentrations (0800–0830 h) during saline infusion in men and women were the lowest in men receiving placebo (P < 0.001) and rose with T treatment (P = 0.003) (Table 1). Bolus-injected mean and peak GH concentrations (0830–1200 h) during saline infusion did not differ by gender or with sex-hormone supplementation. This was also true of the calculated GH distribution volume (GH dose/peak GH concentration). Injected peak GH values averaged 7.2–8.8 μg/liter, representing a nearly physiological GH-feedback signal during saline infusion. Such comparisons are not meaningful on the GHRH and GHRP-2 stimulation days.

Fig. 2.

Plots of mean GH concentration profiles matching the design given in Fig. 1 in 10 men (○) and 10 women (●). Peptide and saline infusions are designated in each panel. Top, Placebo; bottom, sex steroid. The GH pulse beginning at 0830 h was injected to enforce negative feedback. Note different y-axis scales (left to right).

Table 1.

Placebo and sex-hormone treatment

	PL women (n = 10)	E2 women (n = 10)	PL men (n = 10)	T men (n = 10)	P value
Preinjection GH (μg/liter)	0.19 ± 0.12a,b	0.48 ± 0.21a	0.064 ± 0.018b	0.25 ± 0.13a	0.001
Mean injected GH	2.3 ± 1.0	2.0 ± 0.73	1.6 ± 0.13	1.7 ± 0.20	0.54
Peak injected GH	7.8 ± 0.36	7.2 ± 1.5	8.2 ± 0.62	8.8 ± 0.96	0.67
GH distribution volume (liters/kg)	0.053 ± 0.008	0.070 ± 0.016	0.070 ± 0.007	0.064 ± 0.007	0.63
Estradiol (pg/ml)	9.9 ± 0.39a	108 ± 19b	25 ± 1.5c	78 ± 8.1b	<0.001
IGF-I (μg/liter)	166 ± 27a,b	136 ± 19a	215 ± 24a,b	243 ± 26b	0.024
SHBG (nmol/liter)	48 ± 4.0a	89 ± 11b	29 ± 3.2c	23 ± 2.6c	<0.001
T (ng/dl)	ND	ND	381 ± 34a	1099 ± 61b	<0.01

Data are the mean ± sem. E₂ and T were measured by mass spectrometry for this table .ND, Not determined; PL, placebo.

^a–cSuperscript letters were assigned by post hoc Tukey's HSD test. Unshared superscript letters indicate P < 0.05 between groups.

E₂ determined by mass spectrometry on the infusion dates rose to comparable (P > 0.10) concentrations in women and men during sex-steroid supplementation (Table 1). Mass spectrometric T concentrations increased in men administered T (P < 0.01). SHBG and IGF-binding protein-1 were higher and IGF-I concentrations lower in women given E₂ than men given T (each P < 0.025). IGF-binding protein-3 did not differ by gender or treatment.

By three-way ANCOVA using the placebo/saline response as the covariate, gender determined total integrated (P = 0.017) and peak (P = 0.003) GH concentrations during feedback inhibition, but not nadir GH concentrations (P = 0.444) (Supplemental Table 2). Treatment (T/E₂ vs. placebo) augmented all three of the total, nadir, and peak GH recovery in the feedback setting (respectively, P < 0.001, P = 0.018, and P = 0.004). Peptidyl secretagogue effects were also significant (P < 0.001) for all three primary feedback-recovery measures. Each of the three principal ANCOVA models (one each for total, nadir, and peak GH recovery) was significant at P < 0.001 with a strong covariate effect (P < 0.001). This allowed interpretation of two-factor and three-factor interactions (none of the latter was significant). The strongest interactions were for 1) treatment × peptide effects on total and peak GH recovery (both P < 0.001) and 2) gender × treatment (P = 0.017) and gender × peptide (P = 0.012) for nadir GH concentrations (Supplemental Table 2).

Total (integrated) GH concentrations during GH feedback and saline infusion were stimulated severalfold by E₂ (women, P = 0.001 vs. placebo) and weakly by T (men, P = 0.053) (Fig. 3). The effect of E₂ in women exceeded that of both placebo (P < 0.001) and T (P = 0.044) in men. In men and women, GHRH and GHRP-2 (individually P < 0.001) increased total GH recovery compared with saline. Responses to GHRP-2 vs. GHRH did not differ (P = 0.194). There were nonsignificant trends for the GHRH effect in women to exceed that in men (P = 0.063) and for GHRP-2's effect to exceed GHRH's effect in men only (P = 0.056). Notably, neither E₂ nor T further amplified the feedback-attenuating effects of peptidyl secretagogue (both P ≥ 0.95). The treatment-by-peptide interaction was due to the capability of E₂ (P < 0.001) and T (P = 0.053) vs. placebo to augment total GH recovery during saline but not during either GHRH (P > 0.95) or GHRP-2 (P > 0.90) infusion. Thus, total GH recovery during feedback was controlled strongly by E₂, weakly by T, and markedly by each secretagogue.

Fig. 3.

Integrated GH concentrations measured after an iv bolus of GH in men and women during placebo (Pl, left) or sex-steroid (T or E₂, right) administration. Superimposed infusions of saline, GHRH, or GHRP-2 are noted, as described in Fig. 1. Data are the mean ± sem, assessed by three-way ANCOVA. Categorical P values are noted. Overall model P < 0.001 with covariate P < 0.001. The effect of gender was P = 0.017, sex-steroid treatment P < 0.001, and peptide category P < 0.001. Means with unshared (differing) superscript letters are significantly different by the post hoc Tukey's HSD test.

Nadir GH concentrations were stimulated by E₂ in women compared with placebo when assessed across all three infusion types (P = 0.005) but not by T in men (P = 0.999) (Fig. 4). The E₂ effect on nadirs was not significant over placebo in any single session. Peptidyl secretagogue effects on GH nadirs were marked both without and with E₂/T supplementation (P < 0.001 for all four comparisons). The main effect of GHRP-2 on nadir GH concentrations exceeded that of GHRH (P < 0.001), and the latter exceeded that of saline (P < 0.001). GHRP-2's recovery of GH nadirs in women was greater than that of GHRH in men (P = 0.005) but not in women (P = 0.246). GHRP-2 in men was more effective than GHRH in both men (P < 0.001) and women (P = 0.001), which explains a significant gender × peptide interaction (P = 0.012). A gender × treatment (P = 0.017) interaction was explained by greater GHRP-2 stimulation in women and men getting E₂/T than GHRH stimulation in men getting GHRH with (P = 0.015) or without (P = 0.020) T. These effects on GH nadirs contrast with those on total GH recovery.

Fig. 4.

Recovery of nadir GH concentrations during controlled negative feedback. Data are presented as in Fig. 3.

Feedback-inhibited peak GH release was determined by all three of gender, sex steroids, and peptide infusion (each P < 0.005) (Supplemental Table 2). There was a strong treatment × peptide interaction (P < 0.001) (Fig. 5. The main effects of gender were 1) higher peak GH levels in women given E₂ than women given placebo (P = 0.003) and than men given placebo (P < 0.001) during saline infusion and 2) greater peak GH responses to GHRH in women than men assessed across both sex-steroid milieus (P = 0.043). The primary influence of E₂/T treatment was enhanced peak GH recovery in the overall cohorts (P < 0.001 vs. placebo). GHRH and GHRP-2 each markedly increased peak GH in women (P < 0.001) and men (P < 0.001) whether or not E₂/T was administered. Peak GH responses to GHRH and GHRP-2 were similar (P = 0.943) and unaffected by exogenous sex steroids (P > 0.95). An interaction between treatment and peptide was due to E₂'s selective stimulation of peak GH during saline (P = 0.003 vs. placebo) but not GHRH or GHRP-2 infusion (P ≥ 0.99). T did not amplify peak GH recovery during any infusion (P ≥ 0.446 vs. placebo). By post hoc Tukey's HSD test, stimulation by E₂ in women exceeded that by placebo (P < 0.001) but not T (P = 0.856) in men. The effect of T in men during saline infusion was nonsignificant (P = 0.078 trend). Thus, with respect to peak GH recovery from feedback, essential factors were E₂ during saline infusion and GHRH and GHRP-2 drive in either sex-steroid milieu.

Fig. 5.

Recovery of peak GH concentrations in an exogenously imposed GH-feedback paradigm. Data are presented as in Fig. 3.

The timing of nadir and peak GH concentrations did not differ by gender, sex-steroid milieu or secretagogue type (P ≥ 0.36 in men and P ≥ 0.17 in women). Mean nadir GH times (minutes after onset of iv GH injection) in men and women were 215 ± 11 min and 208 ± 9.2 min, respectively (P value nonsignificant). The peak GH times were 372 ± 18 min (men) and 365 ± 17 min (women) (P value nonsignificant).

Exploratory linear-regression analysis revealed that in the placebo setting, E₂ concentrations in the combined cohorts positively predicted GHRP-2-stimulated nadir GH concentrations (R² = 0.30; P = 0.012). In the E₂/T-supplemented milieus, E₂ was a positive correlate of GHRH-stimulated GH nadirs (R² = 0.27; P = 0.018).

Discussion

The present investigation found that estradiol is distinctly more effective than T in promoting GH's recovery from negative feedback, whereas GHRH and GHRP are even more effective. In particular, E₂ and peptide agonists strongly determined all three major model-free measures of feedback-regulated GH secretion, viz., nadir, peak, and mean GH concentrations. Estradiol compared with placebo augmented nadir (by 6-fold) and peak (by 4-fold) GH concentrations markedly during negative feedback. Moreover, E₂ concentrations predicted nadir GH concentrations under GHRP-2 (R² = 0.30) and GHRH (R² = 0.27) drive. The main effect of T supplementation in men was to increase baseline GH concentrations by 4.7-fold (P = 0.001). In both sexes, peptidyl secretagogues more powerfully stimulated integrated, nadir and peak feedback-suppressed GH recovery. These mechanisms could contribute to recognized differences in pulsatile GH secretion in men and women (2, 23).

The study paradigm comprised continuous exogenous peptide stimulation to overcome feedback-inhibited GH secretion. The doses of GHRH and GHRP-2 employed here were chosen because they maintain pulsatile GH secretion for at least 24 h in the absence of exogenous GH feedback (9–11, 24–26). The intention thereby was to obviate possible confounding effects of variable hypothalamo-pituitary availability of the native peptides, GHRH and ghrelin, during peptide infusions. The absolute maximally stimulating doses of continuous iv GHRH and iv GHRP-2 infusions have not been delineated in (older) humans (2). In the two sexes, individual GHRH and GHRP-2 infusions augmented nadir GH concentrations by 11- to 14-fold and by 19- to 53-fold, respectively, over saline. Of the two peptides, GHRP-2 was the much stronger stimulus with respect to both GH nadir responses. E₂ and T administration did not inhibit or potentiate the feedback-rescuing effects of peptidyl secretagogues or alter the greater impact of GHRP-2 than GHRH. A simple interpretation of these outcomes would be that E₂ and T enhance hypothalamo-pituitary sensitivity to (or the potency of) endogenous GHRH and ghrelin in opposing negative feedback during saline infusion. Lack of sex-steroid potentiation of exogenous GHRH action would indicate unchanged efficacy of this peptide in the E₂/T-supplemented feedback context. Although endogenous peptide actions are difficult to quantify directly, E₂ supplementation does enhance exogenous GHRH and GHRP-2 potency and diminish exogenous somatostatin potency but not efficacy (9, 10, 27).

In men only, GHRP-2 enhanced nadir GH concentrations 4.0–4.7-fold more than GHRH infusion in both the placebo and T-supplementation context. This peptide difference was not evident in women. Because castrate levels of T/E₂ were not tested here, further studies will be needed to evaluate whether T or E₂ concentration thresholds exist for this secretagogue distinction. Greater stimulation of nadir GH concentrations by GHRP than GHRH may be due to the capability of the former to antagonize central nervous system (CNS) effects of SS and synergize with GHRH (28–30).

Estrogen supplementation in women elevated feedback-suppressed integrated and peak GH concentrations compared with placebo in the feedback setting. T supplementation in men shared this effect on integrated GH recovery. The actions of E₂/T emerged selectively during saline rather than peptide infusion. Specifically, whereas each peptidyl secretagogue markedly augmented feedback-suppressed integrated and peak GH levels (by 5- to 20-fold over saline-associated levels), the effects of GHRH and GHRP-2 on these endpoints were not modified further by E₂ or T administration. A parsimonious postulate would be that E₂ amplifies endogenous GHRH or ghrelin action, an effect that would be replaced by infusing nearly maximally stimulating amounts of GHRH or GHRP-2. A mechanism underlying this hypothesis may be the capability of E₂ to induce GHRP receptors (31), inasmuch as GHRP/ghrelin seems to oppose CNS actions of SS that restrain GHRH secretion (2, 28). In addition, E₂ can repress CNS GH receptors (32), which are thought to activate SS restraint of arcuate nucleus GHRH neurons (33). The fact that GHRP-R1a knockdown lowers GH and IGF-I concentrations in adult female (but not male) mice would be consistent with these concepts (6).

T tended to elevate total (integrated) (P = 0.053) and peak (P = 0.078) GH recovery under experimentally controlled negative feedback. Larger studies will be needed to assess this point. In a developmental study, midpuberty compared with prepuberty in boys was associated with accelerated GH recovery from negative feedback (34). The 3-fold higher (pharmacological) dose of GH infused, the lack of placebo vs. T injections, and the bolus GHRH design limit direct comparison with the present paradigm of a nearly physiological feedback pulse (injected peak GH = 7.2–8.8 μg/liter) and continuous GHRH and GHRP-2 drive. If the T effect is confirmed in adults, an important question will be whether T acts via the androgen receptor or after aromatization to E₂ in opposing negative feedback (35).

Caveats include the need to confirm outcomes in larger cohorts (here, n = 20 older adults resulting in 120 8-h sampling sessions), extend the dose and duration of E₂ and T supplementation and peptide infusions, evaluate similar mechanisms in a sex steroid-depleted milieu, and ultimately relate feedback regulation to age, body-compositional, and other variables. Further investigations are needed to clarify the modulatory roles of growth-hormone binding protein (high concentrations would in principle extend the GH half-life) (36, 37) and of IGF-I and its binding proteins in GH autofeedback and E₂/T effects (38). The clinical paradigm presented here may have utility in such studies.

In conclusion, healthy older women and men manifest distinct GH responses to E₂ and T supplementation and individual GHRH and GHRP stimulation during exogenously controlled negative feedback. Distinctions apply to the key pulsatility variables, nadir (maximally suppressed) and peak (maximally escaped) GH concentrations. Accordingly, at least three factors, gender, sex-steroid milieu, and secretagogue availability, must be considered in evaluating feedback-regulated GH output in healthy older individuals.