Previous studies have established that an increase of serum S-100 indicates damage to glial cells and blood-brain barrier dysfunction—confirming head injury in patients. Building upon that finding, this study set out to assess if S-100 levels in head injury patients served as a marker for determining the severity of the injury and possible neuronal damage. Of the 41 patients involved in the study, 30 had severe head injury, and 11 minor. After taking the blood samples from the patients (less than 6 hours following injury) and running statistical tests, the authors found a significant difference between the two groups. 27 (66%) had an elevated serum level of S-100, and 25 of those patients had sustained the severe head injury. Furthermore, the mean serum level in the severe head injury group was 2.6, whereas in the minor group it only reached .35. The S-100 serum level seems to reflect the severity of central nervous system trauma.

Type and Number of Cases      Range of S-100 Serum Levels

Minor Head Injury (11)

.2 to .8

Severe Head Injury (30)

.35 to 16.2

Severe Head Injury with
Favorable Outcome (21)

.35 to 10

Severe Head Injury with
Unfavorable Outcome (9)

.7 to 16.2

The authors emphasize the need for early testing of S-100 serum levels, since they rapidly increase. Yet, they point out:

"[T]he early outcome estimation does not consider rehabilitation measurements and is therefore probably more reliable when deciding which group of patients needs more intensive monitoring in specialized centers in the acute posttraumatic phase. Our goal in this study was not to describe a single serum marker that is able to predict outcome after a longer period. Instead, we would like to see the S-100 serum level not as a sole determinant in outcome prediction but as a helpful tool, in addition to clinical or radiologic data, for predicting survival after head injury."

Rothoerl R, Woertgen C, Holzschuh M, et al. S-100 serum levels after minor and major head injury. The Journal of Trauma: Injury, Infection, and Critical Care 1998;45(4):765-767.
" "