Previous experimental data have demonstrated that steroid pretreatment regulates endotoxin-elicited cytokine production. The timing of such hypercortisolemia also appears to be a major determinant of both the systemic and the cytokine response to infectious stimuli. Our study was undertaken to further study the in vivo influence of glucocorticoid infusion concurrent with and before endotoxin exposure in man. A total of 23 normal human subjects were given endotoxin (LPS) alone or pretreated with hydrocortisone infusion for 6 h immediately before and concomitant to LPS administration; or rendered hypercortisolemic for a 6-h period waiting 6, 12, or 144 h before LPS administration. LPS administration was followed by significant elevations in temperature (1.9 +/- 0.3 degrees C), pulse (29 +/- 6 bpm), and resting energy expenditure (26.2 +/- 0.4 kcal/kg/day) as well as epinephrine (236 +/- 59 pg/ml), cortisol (296 +/- 29), and C-reactive protein (4.2 +/- .03 mg/dl) as compared with base-line values. Levels of TNF and IL-6 that were not detectable before LPS administration, peaked, respectively, at 90 and 120 min after LPS (155 +/- 4 pg/ml, 12 +/- 1 U/ml). Glucocorticoids when given immediately before and concomitant with LPS significantly attenuated the temperature (0.8 +/- 0.01 degrees C) and pulse rate response (10 +/- 3 bpm) seen after LPS alone as well as suppressing peak levels of epinephrine (78 +/- 14 pg/ml) and C-reactive protein (undetected). TNF remained undetectable in this group although the IL-6 response (14 +/- 1 U/ml) was unchanged. With a 6-h interval between hydrocortisone infusion and LPS challenge, changes in temperature, pulse, resting energy expenditure, and hormone levels were similar to those seen after LPS alone whereas TNF remained undetectable and IL-6 levels were similar to subjects receiving LPS alone. Subjects receiving LPS after 12 or 144 h after hydrocortisone infusion displayed hemodynamic and hormonal responses similar to the LPS alone group, yet mounted significantly greater circulating levels of both IL-6 (117 +/- 14 U/ml, 160 +/- 51 U/ml at 12 and 144 h) and TNF (609 +/- 173, 671 +/- 132 pg/ml at 12 and 144 h) to those observed after LPS alone. We conclude that antecedent periods of hypercortisolemia participate in regulation of the hemodynamic, hormonal, and cytokine responses to endotoxin and that a complex temporal relationship between hypercortisolemia and LPS induced cytokine and systemic responses exists.