Fasting as a metabolic stress paradigm selectively amplifies cortisol secretory burst mass and delays the time of maximal nyctohemeral cortisol concentrations in healthy men

Serum cortisol concentrations are increased in fasted or malnourished human subjects. The dynamic mechanisms underlying this adaptive response have been investigated in eight normal men by analyzing serum cortisol concentrations measured in blood obtained at 5-min intervals over 24 h on a control (fed) day and on the fifth day of a fast (water only) assigned in randomized order. A multiple parameter deconvolution method was used to simultaneously resolve endogenous cortisol secretion and half-life. Five days of fasting induced a 1.8-fold increase in the 24-h endogenous cortisol production rate (fed, 2504 +/- 308; fasted, 4528 +/- 488 nmol/L distribution volume; P < 0.006). This enhanced cortisol production rate was accounted for by a 1.6-fold increase in the mass of cortisol secreted per burst (fed, 115 +/- 12.1; fasted, 183 +/- 17.3 nmol/L; P < 0.02). Cortisol secretory event amplitudes (maximal rates of cortisol release attained within a burst) increased in seven of eight men, and mean secretory burst durations remained unchanged by fasting. Moreover, the number of computer-resolved cortisol secretory bursts per 24 h (fed, 22 +/- 1.4; fasted, 25 +/- 2.0; P = NS) and the interburst interval (fed, 65 +/- 4.0; fasted, 57 +/- 4.4 min) did not change significantly during a 5-day fast. The calculated half-life of endogenous cortisol was not significantly altered by fasting (fed, 108 +/- 9.7; fasted, 129 +/- 11 min). There was no significant change in the nyctohemeral pattern of varying adrenocortical secretory burst frequency in response to fasting. However, the mean (mesor) mass of glucocorticoid secreted per burst over 24 h rose significantly in response to fasting. In addition, by cosinor analysis, maximal serum cortisol concentrations occurred (95% confidence intervals) between 0930-1334 h in the fed state and between 1116-1612 h in the fasted state (P < 0.04). Fasting augmented the mesor (average value about which the diurnal rhythm oscillates; P < 0.0008 compared with fed state) and the amplitude (P < 0.04) of the 24-h serum cortisol concentration profile. Linear regression analysis disclosed a significant inverse relationship between mean serum cortisol and GH concentrations in fasted men (r = -0.76; P < 0.02). In conclusion, the present data indicate that starvation-induced enhancement of cortisol secretion in young healthy men is mediated by an increased glucocorticoid secretory burst mass, rather than changes in secretory burst frequency or duration or in cortisol half-life. In addition, fasting modifies the diurnal secretory pattern of cortisol by delaying maximal serum concentrations to the early afternoon. The inverse relationship between serum cortisol and GH responses to fasting suggests differential regulation of the corticotropic and somatotropic axis by the metabolic stress of fasting and/or feedback interactions between these two axes when they are both activated.