An ensemble model of the male gonadal axis: illustrative application in aging men.

Testosterone (Te) production declines in the aging male, albeit for unknown reasons. Plausible mechanisms include reduced secretion of GnRH, less feedforward by LH, and/or altered feedback by systemic Te. The present study tests all three postulates in a cohort of 10 young (20-35 yr old) and eight older (50-72 yr old) men. The experimental paradigm comprised graded blockade of the GnRH receptor to create four distinct strata of LH and Te pulsatility in each subject. A novel analytical formalism was developed to reconstruct implicit dose-response functions linking 1) virtual GnRH outflow positively to LH secretion, 2) LH pulses positively to Te secretion, and 3) Te concentrations negatively to the size and number of LH secretory bursts. Validation was by direct pituitary sampling in the horse and sheep. Statistical comparisons disclosed that age decreased the efficacy of each of 1) virtual GnRH outflow (P < 0.01), 2) LH drive of Te secretion (P < 0.01), and 3) total, bioavailable and free Te feedback on GnRH-driven LH secretion (P = 0.015). In contrast, age increased the potency of virtual GnRH feedforward (P = 0.013) and did not affect Te's inhibition of LH pulse frequency. Unexplained variance was less than 10%. Robustness was shown by resampling techniques. Accordingly, competitive silencing of one locus of control and ensemble-based analyses identified triple regulatory deficits in the aging male gonadal axis. The generality of the present integrative model suggests utility in parsing interlinked adaptations in other physiological networks.