The adrenocortical response to brain injury: correlation with the severity of neurologic dysfunction, effects of intoxication, and patient outcome.

To test the hypothesis that cortisol levels reflect the extent of neurologic dysfunction and predict patient outcome, neurologic function and cortisol levels were determined in 120 traumatically brain injured patients who never received glucocorticoid treatment. Their mean age was 29 years and 78% were men. The impact of intoxication was examined in 59 patients who had ethanol levels measured. Ethanol was detectable in 40 patients and greater than or equal to 100 mg/dl in 31. There were significant correlations between the extent of neurologic dysfunction, determined by the Glasgow Coma Score and plasma cortisol concentrations 1 and 4 days postaccident. Cortisol levels were universally elevated on admission and approached normal 7 days later. Multiple linear regression analysis revealed significant effects of circulating ethanol levels on the association between cortisol concentrations and progressively worsening neurologic function, i.e., ethanol reduced the magnitude of the cortisol elevations in a dose dependent manner, abolishing this relationship at levels above 100 mg/dl. Analysis of the relationships between circulating cortisol levels and patient outcome provided a second method for ascertaining the association between injury severity and the magnitude of adrenocortical activation. Admission and day 1 cortisol concentrations were 25 to 40% lower in patients having good recoveries or moderate disabilities than those who remained severely disabled, persistently vegetative or died; serum cortisol values of less than 20 micrograms/dl one day after the accident were more likely to be associated with a good outcome than a poor one (55 vs. 25%, p less than 0.001). The worsening prognosis of patients having higher cortisol values is further reflected in the duration of acute hospitalization of these individuals.(ABSTRACT TRUNCATED AT 250 WORDS)