Anxiety disorders are common in patients with TBI and range in frequency from 11 to 70% (Klonoff, 1971; Lewis, 1942). In their review of 12 studies conducted between 1942 and 1990, Epstein and Ursano (1994) noted that in a total sample of 1,199 subjects, most of whom had featured only minor head injuries, 29% of postinjury subjects featured anxiety disorders.
Deb, Lyons, Koutzoukis, Ali et al. (1999) followed up 196 patients who were admitted to the hospital between July 1, 1994 and June 30, 1995 with a traumatic brain injury. Each of these patients had a period a loss of consciousness, radiological evidence of brain assault, or a GCS of less than 15. Of the 120 patients interviewed between the ages of 18 and 64, 26 (21.7%) had an International Classification of Diseases (ICD)-10 diagnosis of psychiatric illness. Of this group, 17 patients (14%) had diagnoses of anxiety disorder including 11 (9%) with panic disorder, 3 (2.5%) with generalised anxiety disorder, 1 (0.8%) with phobic disorder, and 2 (1.6%) with OCD.
Despite the likely association between TBI and the anxiety disorders there is still some inconsistency in this literature. One of the most notable inconsistencies is in the dose response relationships of the presence and severity of the anxiety disorders relative to the severity of injury. Sir Austin Bradford Hill first proposed the concept of the "biologic gradient" in 1965. He proposed that one of the criteria for a causative rather than a correlational relationship between a particular clinical entity (in this case post-TBI anxiety disorder) and the putative cause (i.e., the injury) was the degree to which the severity of the causative entity correlated with the severity of the presentation. This effect is often reported to be inconsistent across studies and in many cases paradoxically, the relationship operates in a manner featuring a negative gradient, with often less severe injuries (i.e., MTBI) resulting in more anxiety symptoms than either moderate or severe TBI (Rapoport, McCauley, Levin, Song, & Feinstein, 2002; van Reekum, Cohen, & Wong, 2000). Clearly any direct association between the two phenomena cannot easily explain this type of data and many factors that mediate the relationship between the two conditions, including the PCS discussed in the last chapter, must significantly colour the expression of the pathology.
Was this article helpful?