Estrogen and Testosterone, But Not a Nonaromatizable Androgen, Direct Network Integration of the Hypothalamo-Somatotrope (Growth Hormone)-Insulin-Like Growth Factor I Axis in the Human: Evidence from Pubertal Pathophysiology and Sex-Steroid Hormone Replacement
Activation of the gonadotropic and somatotropic axes in puberty is marked by striking amplification of pulsatile neurohormone secretion. In addition, each axis, as a whole, constitutes a regulated network whose feedback relationships are likely to manifest important changes at the time of puberty. Here, we use the regularity statistic, approximate en- tropy (ApEn), to assess feedback activity within the somatotropic (hy- pothalamo-pituitary/GH-insulin-like growth factor I) axis indirectly. To this end, we studied pubertal boys and prepubertal girls or boys with sex-steroid hormone deficiency treated short-term with estrogen, tes- tosterone, or a nonaromatizable androgen in a total of 3 paradigms. First, our cross-sectional analysis of 53 boys at various stages of puberty or young adulthood revealed that mean ApEn, taken as a measure of feed- back complexity, of 24-h serum GH concentration profiles is maximal in pre- and mid-late puberty, followed by a significant decline in postpu- bertal adolescence and young adulthood (P 5 0.0008 by ANOVA). This indicates that marked disorderliness of the GH release process occurs in mid-late puberty at or near the time of peak growth velocity, with a return to maximal orderliness thereafter at reproductive maturity. Sec- ond, oral administration of ethinyl estradiol for 5 weeks to 7 prepubertal girls with Turner's syndrome also augmented ApEn significantly (P 5 0.018), thus showing that estrogen per se can induce greater irregularity of GH secretion. Third, in 5 boys with constitutionally delayed puberty, im testosterone administration also significantly increased ApEn of 24-h GH time series (P 5 0.0045). In counterpoint, 5 a-dihydrotestosterone, a nonaromatizable androgen, failed to produce a significant ApEn in- crease (P. 0.43). We conclude from these three distinct experimental contexts that aromatization of testosterone to estrogen in boys, or es- trogen itself in girls, is likely the proximate sex-steroid stimulus am- plifying secretory activity of the GH axis in puberty. In addition, based on inferences derived from mathematical models that mechanistically link increased disorderliness (higher ApEn) to network changes, we suggest that sex-steroid hormones in normal puberty modulate feedback within, and hence network function of, the hypothalamo-pituitary/GH- insulin-like growth factor I axis. (J Clin Endocrinol Metab 82: 3414- 3420, 1997)