Objective: Activation and suppression of immune responses are crucial events during sepsis. Based on substantial new data, a complex picture of differential immune-enhancing and immunosuppressive actions of adrenocortical steroids is emerging. The adrenal androgen dehydroepiandrosterone and its precursor, dehydroepiandrosterone-sulfate, show a considerable decrease with increasing age and serve as functional antagonists to endogenous glucocorticoids. Therefore, we examined time-dependent changes in dehydroepiandrosterone, dehydroepiandrosterone-sulfate, cortisol, adrenocorticotropin, and inflammatory variables in surviving and nonsurviving patients with severe sepsis.
Design: Prospective observational study in consecutive patients.
Setting: Medical and interdisciplinary intensive care units in two university hospitals and one city hospital.
Patients: Thirty nonsurgical patients (25 men and 5 women) with severe sepsis (American College of Chest Physicians/Society of Critical Care Medicine criteria); 15 survivors (mean age, 54 +/- 14 yrs; Acute Physiology and Chronic Health Evaluation III score, 59 +/- 35) and 15 nonsurvivors (mean age, 63 +/- 15 yrs; Acute Physiology and Chronic Health Evaluation III score, 67 +/- 24) were included. Hormones were compared individually and between survivors/nonsurvivors by sequential blood drawings from early sepsis till time of recovery/death.
Interventions: None.
Measurements and main results: During early sepsis, cortisol (nmol/L) was not significantly higher in survivors than nonsurvivors (750 +/- 121 vs. 454 +/- 92, p <.08) and decreased in survivors (p <.01) during late sepsis. During early sepsis, dehydroepiandrosterone-sulfate (percentage of age-matched normal levels) was higher in survivors than nonsurvivors (85 +/- 19 vs. 22 +/- 7, p <.01). Dehydroepiandrosterone-sulfate decreased in survivors (p =.0001) but remained low in nonsurvivors during late sepsis. Dehydroepiandrosterone (percentage of age-matched normal levels) was not significantly elevated in survivors compared to nonsurvivors during early sepsis (282 +/- 42 vs. 214 +/- 63, p <.08). Dehydroepiandrosterone decreased in survivors (p <.01) but not in nonsurvivors during late sepsis. Linear regression for dehydroepiandrosterone levels showed a reconstitution of age dependence only in survivors during recovery. Adrenocorticotropin levels did not change. The dehydroepiandrosterone-sulfate/cortisol ratio decreased significantly in both survivors and nonsurvivors, whereas dehydroepiandrosterone/cortisol ratio only decreased in survivors during course of sepsis.
Conclusions: During sepsis, adrenal androgens and glucocorticoids show a diverse time-dependent course in survivors and nonsurvivors.