Short-term estradiol replacement in postmenopausal women selectively mutes somatostatin's dose-dependent inhibition of fasting growth hormone secretion.

How estradiol stimulates pulsatile GH secretion in the human is not well understood. Here, we test the clinical hypothesis that estradiol stimulates GH secretion, in part, by opposing somatostatin's inhibition of GH release. To this end, 13 estrogen-withdrawn postmenopausal women received placebo or 1 mg micronized estradiol-17beta orally, twice daily for 14 days, in a prospectively randomized, patient-blinded, within-subject cross-over design. For each intervention, the dose-dependent suppressive actions of somatostatin were evaluated by infusing 0 (saline), 3, 10, 30, 100, or 300 microg/1.73 m(2).h somatostatin-14 continuously, iv, for 3 h, on separate mornings, in the fasting state, 48 h apart. Blood was sampled at 10-min intervals for 2 h before, for 3 h concurrently with, and for 1 h after each infusion. Serum GH concentrations were quantitated in an ultrasensitive chemiluminescence-based assay (detection threshold, 0.005 microg/L). In the estrogen-deficient milieu, constant iv somatostatin infusions inhibited steady-state serum GH concentrations (valley mean during the last 60 min of the infusion interval) in a dose-dependent manner (P < 10(-4) interventional effect). Maximally effective doses of somatostatin reduced the latter by 89 +/- 6.1% (mean +/- SEM) below the subject-specific preinfusion baseline. Estrogen administration increased the serum estradiol concentration from 12 +/- 1 to 245 +/- 35 pg/mL [42 +/- 4 to 920 +/- 110 pmol/L] (P < 10(-4)); decreased serum concentrations of LH (P = 0.018), FSH (P < 10(-4)), and insulin-like growth factor-I (P = 0.003); and elevated the fasting (6-h mean) serum GH concentration from 0.41 +/- 0.07 to 0.87 +/- 0.27 (P = 0.011). Estradiol supplementation did not alter somatostatin's maximal suppression of GH by 89 +/- 4.7% (P < 10(-4) below subject-specific preinfusion baseline), thus signifying unchanging somatostatin efficacy. In contrast, estradiol replacement significantly elevated the half-maximally inhibitory dose of infused somatostatin by 13.5-fold, from 0.43 (0.38-0.48, 95% group statistical confidence intervals) (placebo) to 6.0 (5.2-7.0) (estradiol) microg/1.73 m(2)/h (P < 10(-4)), denoting muting of somatostatin's inhibitory potency. The latter inference was confirmed by a concomitant 4-fold decrease in the exponential steepness of the somatostatin inhibitory dose-response function; viz., mean 1.42 (1.49 to 1.33) (placebo) vs. 0.34 (0.62 to 0.26) (estradiol) slope units (P < 10(-4)). The foregoing effects were specific, because estrogen did not alter somatostatin's dose-dependent enhancement (P < 10(-4)) of the orderliness of GH release patterns, as quantitated via the approximate entropy regularity statistic. In summary, short-term replacement of estradiol to midfollicular phase levels in postmenopausal women selectively reduces the potency, but not the efficacy, of somatostatin's dose-dependent inhibition of GH release. Estrogen supplementation does not modify somatostatin's reciprocal enhancement of the quantifiable orderliness (approximate entropy) of the GH secretory process. Accordingly, we postulate that estradiol can facilitate pulsatile GH secretion, in part, by opposing the repressive actions of somatostatin.