Circadian, ultradian, and episodic release of beta-endorphin in men, and its temporal coupling with cortisol.

beta-Endorphin and ACTH derive from a common peptide precursor. Although much is known about the physiological patterns of ACTH release, neither the minute to minute regulation of beta-endorphin secretion nor its temporal relationship to cortisol has been characterized. As an initial step to defining the regulation of beta-endorphin release in man, we studied the circadian periodicity, ultradian rhythmicity, and episodic pulsatility of serum beta-endorphin concentrations in seven normal men. Blood sampling was conducted at 10-min intervals for 24 h, and the subsequent serum samples were assayed by a two-site immunoradiometric assay. Computerized analysis of the subsequent beta-endorphin time series revealed a mean beta-endorphin pulse frequency of 13 +/- 1 (+/- SE) peaks/24 h, corresponding to an interpulse interval of 100 +/- 7 min. The mean maximal peak height of beta-endorphin pulses was 31 +/- 3 pg/mL (9.0 +/- 0.8 pmol/L), which represented an incremental increase of 11 +/- 1 pg/mL 3.2 +/- 0.4 pmol/L; 63 +/- 13%) above the preceding nadir. The average beta-endorphin peak exhibited a duration of 68 +/- 6 min. Fourier analysis revealed a significant circadian amplitude of 6 +/- 1 pg/mL (1.6 +/- 0.4 pmol/L; 23% of the 24-h mean concentration), with an acrophase (time of maximum value) at 1043 h (+/- 40 min). Spectral analysis also disclosed beta-endorphin rhythms with mean periodicities of 29 +/- 4, 42 +/- 4, and 61 +/- 5 min. Gel filtration chromatography confirmed that serum beta-endorphin peaks contained significantly more immunoactive beta-endorphin [62 pg/mL (18 pmol/L)] than did the flanking nadirs [16 and 18 pg/mL (4.6 and 5.2 pmol/L)]. Auto- and cross-correlation analyses of serum beta-endorphin and cortisol concentrations followed by autoregressive modeling disclosed that all seven men had significant positive cross-correlations between serum beta-endorphin and cortisol considered simultaneously or when cortisol lagged beta-endorphin by 10 min. A negative cross-correlation was found in five of the seven men when cortisol was considered to lead beta-endorphin by 20 or 30 minutes. We conclude that beta-endorphin is released physiologically in a pulsatile manner with circadian and ultradian rhythmicity and a close temporal coupling to cortisol.