Pulsatile hormone secretion during severe sepsis: Accuracy of different blood sampling regimens

The metabolic response to sepsis is dependent on the hormonal status. However, reported plasma hormone levels vary widely among studies. The persistence of pulsatile secretion, as occurs normally, may explain the observed variability. To study whether pulsatile hormone secretion persists during sepsis and how it affects assessment of the hormonal status from single measurements, we measured growth hormone (GH), prolactin, cortisol, insulin, and C-peptide at 20-minute intervals for 24 hours in eight consecutive patients with severe sepsis. Twenty-four-hour averages (mean ± SD) were 3.3 ± 2.5 ng/mL for GH, 640 ± 461 nmol/L for cortisol, 18.2 ± 4.8 mU/L for insulin, and 3.4 ± 2.9 U/L for C-peptide, at a pulse frequency between 3.3 ± 2.7 for C-peptide and 10.2 ± 3.4 for insulin, and an increase of the maximal value in a pulse above the preceding nadir of 131% ± 13% for cortisol and 376% ± 386% for GH, as assessed with Cluster analysis. Prolactin levels were below the detection limit in all but one patient, probably due to the administration of dopamine. To determine the accuracy of less frequent blood sampling regimens, we simulated different sampling strategies and compared them with the 24-hour averages. The accuracy of single samples proved inadequate for all hormones. Sampling every 20 minutes for periods of 4, 8, or 12 hours improved accuracy, but intermittent sampling every 1, 2, 4, or 6 hours during a 24-hour period yielded even more accurate results. Measuring GH every hour, insulin every 4 hours, and cortisol and C-peptide every 6 hours would represent the minimum number of samples necessary to obtain a result deviating less than 10% from the 24-hour average. In view of persistent pulsatile secretion, the latter regimens are advocated for proper assessment of the hormonal status during sepsis.