Background. In trauma patients a specific pattern of the hormonoendocr
ine response to injury has been described. Conflicting data exist rega
rding the hormonal changes during sepsis. We studied the time course a
nd variability of the hormonal and metabolic responses to severe sepsi
s in twelve medical patients. Methods. Hormone levels were measured ev
ery 6 hours. Urinary nitrogen and 3-methylhistidine excretion were mea
sured daily. Indirect calorimetry was used to calculate energy product
ion rate (EPR). Results. None of the hormones measured showed a time c
ourse in individual patients. Growth hormone (1.69 +/- 0.17 ng/ml) and
insulin (48 +/- 5 mU/L) values were within normal range. Cortisol val
ue (474 +/- 44 nmol/L) was in the upper normal range. Insulin-like gro
wth factor 1 level (12.5 +/- 5.7 nmol/L) was decreased. Nitrogen balan
ce was negative and did not change during the study period. Neither ox
ygen consumption (VO2) (318 +/- 20 ml/min), carbon dioxide output (283
+/- 10 ml/min), EPR (2064 +/- 435 kcal/day), nor fuel utilization sho
wed a time course. The variability during the study period compared wi
th values on admission ranged from 20% (insulin-like growth factor 1)
to 74% (growth hormone). The variability of calculated fuel utilizatio
n ranged from 35% (protein) to 180% (fat). The variability of VO2, car
bon dioxide production, and EPR was far less, 8%. Conclusions. A speci
fic pattern in the endocrine and metabolic responses as observed in tr
auma patients does not occur in medical patients with severe sepsis. H
ormones and calculated fuel utilization show large variations when com
pared with values on admission. VO2 and EPR vary far less. Nutritional
needs in patients with sepsis, therefore, can be calculated on admiss
ion and need not be recalculated during the first week after admission
.