Pathophysiology of hypercortisolism in depression.

OBJECTIVE: The mechanisms mediating hypercortisolemia in depression remain controversial. Adopting the biomarker strategy, we studied adrenocorticotropin (ACTH) and cortisol dynamics in hypercortisolemic and non-hypercortisolemic depressed in-patients, and in normal volunteers. METHOD: Deconvolution analysis of 24-h pulsatile secretion, approximate entropy (ApEn) estimation of secretory regularity, cross-ApEn quantitation of forward and reverse ACTH-cortisol synchrony, and cosine regression of 24-h rhythmicity. RESULTS: Hypercortisolemia was strongly associated with melancholic and psychotic depressive subtypes. Hypercortisolemic patients had elevated ACTH and cortisol secretion, mediated chiefly by increased burst masses. Basal ACTH secretion was increased, ACTH half-life was reduced, and mean 24-h ACTH concentration was normal. Cortisol secretion was increased in a highly irregular pattern (high ApEn), with high ACTH --> cortisol cross-ApEn (impaired feedforward coupling). Cortisol-mediated feedback on the secretory pattern of ACTH was normal. Hypercortisolemic depressed patients had normal programming of the central hypothalamo-pituitary-adrenal (HPA) axis pulse generator: ACTH pulse frequency, cortisol pulse frequency, circadian acrophases, and ApEn of ACTH secretion were normal. Responsiveness of the adrenal cortex to endogenous ACTH was normal. Non-hypercortisolemic patients resembled hypercortisolemic patients on ACTH regulatory parameters but had low total cortisol secretion. CONCLUSION: Increased ACTH secretion occurs in depressed in-patients regardless of cortisolemic status, confirming central HPA axis overdrive in severe depression. Depressive hypercortisolemia results from an additional change in the adrenal cortex that causes ACTH-independent, disorderly basal cortisol release, a sign of physiological stress in melancholic/psychotic depression.

Corticotropin secretory dynamics in humans under low glucocorticoid feedback.

To explore the mechanisms of homeostatic adaptation of the hypothalamo-pituitary-adrenal axis to an experimental low-feedback condition, we quantitated pulsatile (ultradian), entropic (pattern-sensitive), and 24-h rhythmic (circadian) ACTH secretion during high-dose metyrapone blockade (2 g orally every 2 h for 12 h, and then 1 g every 2 h for 12 h). Plasma ACTH and cortisol concentrations were sampled concurrently every 10 min for 24 h in nine adults. The metyrapone regimen reduced the amplitude of nyctohemeral cortisol rhythm by 45% (P = 0.0013) and delayed the time of the cortisol maximum (acrophase) by 7.1 h (P = 0.0002). Attenuated cortisol negative feedback stimulated a 7-fold increase in the mean (24-h) plasma ACTH concentration, which rose from 24 +/- 1.6 to 169 +/- 31 pg/ml (ng/liter) (P < 0.0001). Augmented ACTH output was driven by a 12-fold amplification of ACTH secretory burst mass (integral of the underlying secretory pulse) (21 +/- 3.1 to 255 +/- 64 pg/ml; P < 0.0001), yielding a higher percentage of ACTH secreted in pulses (53 +/- 3.5 vs. 92 +/- 1.3%; P < 0.0001). There were minimal elevations in basal (nonpulsatile) ACTH secretion (by 50%; P = 0.0049) and ACTH secretory burst frequency (by 36%; P = 0.031). The estimated half-life of ACTH (median, 22 min) and the calculated ACTH secretory burst half-duration (pulse event duration at half-maximal amplitude) (median, 23 min) did not change. Hypocortisolemia evoked remarkably more orderly subordinate patterns of serial ACTH release, as quantitated by the approximate entropy statistic (P = 0.003). This finding was explained by enhanced regularity of successive ACTH secretory pulse mass values (P = 0.032). In contrast, there was no alteration in serial ACTH interpulse-interval (waiting-time) regularity. At the level of 24-h ACTH rhythmicity, cortisol withdrawal enhanced the daily rhythm in ACTH secretory burst mass by 29-fold, elevated the mesor by 16-fold, and delayed the acrophase by 3.4 h from 0831 h to 1154 h (each P < 10(-3)). In summary, short-term glucocorticoid feedback deprivation primarily (>97% of effect) amplifies pulsatile ACTH secretory burst mass, while minimally elevating basal/nonpulsatile ACTH secretion and ACTH pulse frequency. Reduced cortisol feedback paradoxically elicits more orderly (less entropic) patterns of ACTH release due to emergence of more regular ACTH pulse mass sequences. Cortisol withdrawal concurrently heightens the amplitude and mesor of 24-h rhythmic ACTH release and delays the timing of the ACTH acrophase. In contrast, the duration of underlying ACTH secretory episodes is not affected, which indicates that normal pulse termination may be programmed centrally rather than imposed by rapid negative feedback. Accordingly, we hypothesize that adrenal glucocorticoid negative feedback controls hypothalamo-pituitary-adrenal axis dynamics via the 3-fold distinct mechanisms of repressing the mass of ACTH secretory bursts, reducing the orderliness of the corticotrope release process, and modulating the intrinsic diurnal rhythmicity of the hypothalamo-corticotrope unit.