218 TEXTBOOK OF TRAUMATIC BRAIN INJURY logical deficits, with lower IQ, worse verbal and visual memory, and language impairment. It could not be deter- mined, of course, whether these factors preceded or re- sulted from the TBI. Strengths of this study include match- ing of age and gender in control subjects, use of operationalized criteria for TBI and “schizophrenia-like psychosis following TBI,” consistent ascertainment of cases and control subjects, direct patient interviews and use of informants, and collection of both structural imaging and neuropsychological data (although neuroimaging data were qualitative and read by different radiologists and a standard neuropsychological battery was not used). What Predicts Psychosis in Brain-Injured Individuals? The preceding studies are the most recent and perhaps most methodologically sound attempts at clarifying the characteristics of injury that place someone at risk for developing psychosis after brain injury. A variety of other studies have looked at other specific factors that may contribute to the development of posttraumatic psycho- sis, including location and extent of injury, and genetic vulnerability. Location of Injury Accumulated evidence suggests that injuries to the left hemisphere and to the temporal lobes may be most closely associated with risk of posttraumatic psychosis (Davison and Bagley 1969). As noted, Sachdev et al. (2001) found that those with a TBI who developed psychosis had more CT scan evidence of brain damage, especially in the left temporal and parietal regions, than those who did not develop a psychosis, though this did not survive Bonferroni correction. In a logistic regression model, only left tempo- ral damage significantly predicted the occurrence of psy- chosis after TBI. In an earlier study, Hillbom (1960) found that 40% of individuals with posttraumatic psychosis had temporal lobe injuries, a significantly higher occurrence than in those with nonpsychotic psychiatric disturbance. Of the group with psychosis, 63% had left-hemisphere injuries (a higher value than for nonpsychotic psychiatric disturbance), 26% had right-hemisphere lesions, and 11% had bilateral injuries. The individuals with schizophrenia- like syndromes had more severe injuries and were more likely to have left hemispheric injury. Koufen and Hagel (1987) evaluated electroencephalo- graphic abnormalities in a cohort of 100 patients with psychosis on a brain injury hospital ward and found that posttraumatic psychosis was associated with abnormal foci in the temporal lobes bilaterally in the majority of cases. However, in this study, psychosis was not well de- fined, and criteria for the diagnosis of posttraumatic psy- chosis were not well described. The suggestion of a link between left-hemisphere in- jury, particularly of the temporal lobe, and psychosis is consistent with findings in other neurological disorders. Davison and Bagley (1969) found that in a series of 150 cases of schizophrenia-like psychoses related to diverse neurological disorders, the lesions were usually in the left hemisphere and temporal lobes. Severity of Injury Many studies have found that severity of TBI is related to risk of posttraumatic psychosis. As early as the 1960s, Davi- son and Bagley (1969) found in their review of eight studies that increased severity of injury with more diffuse brain damage and coma longer than 24 hours were risk factors for the development of posttraumatic psychosis. Thomsen (1984) also found a link between severity of brain injury and subsequent psychosis. Hillbom (1960) found that the rate of psychosis increased with the severity of the injury: 2.8% of those with mild injuries, 7.2% of those with medium-severity injuries, and 14.8% of those with severe injuries had become psychotic. Furthermore, in Hillbom’s study, the patients who appeared to have schizophrenia had more severe injuries than the other patients with psychosis. These findings are corroborated by the more rigorous case-control study of Sachdev et al. (2001), who found that measures of injury severity, including duration of uncon- sciousness, evidence of brain damage on CT scan, and cog- nitive deficits on neuropsychological testing, predicted posttraumatic schizophrenia-like psychosis. However, the link between injury severity and psy- chosis is not a universal finding. Violon and De Mol (1987) found that severity of injury did not predict psy- chosis after TBI. In the Fujii and Ahmed (2001) study noted earlier, there was a trend for the control group to have had more severe injuries. In the posttraumatic psy- chosis group, 16 of 22 patients had only had a mild brain injury. Also, for members of families with a history of bi- polar disorder and schizophrenia, the risk of developing schizophrenia associated with having had a TBI was found to be unrelated to the severity of the TBI (Malaspina et al. 2001). Other Features of Injury The type of brain injury may also be related to psychosis risk. Davison and Bagley (1969) found that closed-head Psychotic Disorders 219 injury was related to risk of posttraumatic psychosis, and Lishman (1968) found a low rate of psychosis after pene- trating head injury in veterans (though follow-up was only 4 years). However, newer studies have not found a link between psychosis risk and type of injury (closed vs. open) (Fujii and Ahmed 2001; Sachdev et al. 2001). Age at injury has not been found to determine psychosis risk (Fujii and Ahmed 2001); nor have behavioral and person- ality changes after TBI (Sachdev et al. 2001). Inherent Vulnerability to Psychosis Risk of posttraumatic psychosis has been linked to pre- traumatic psychological characteristics and vulnerability to psychosis. Previous psychopathological disturbances have been reported for 83% of individuals who develop psychosis after TBI (Violon and De Mol 1987). Lishman (1987) found that psychosis is more likely to follow TBI in individuals who are predisposed to schizophrenia. In the recent study by Sachdev et al. (2001) genetic vulnera- bility to psychosis, as indicated by having a first-degree relative with a psychotic disorder, was found to be among the strongest predictors of who would develop psychosis after a TBI. Gender There are no studies that clearly evaluate the role of gen- der in risk for posttraumatic psychosis. Many of the ear- lier studies focused on veterans, who were invariably men. Although Fujii and Ahmed (2001) reported a preponder- ance of males in a sample of state hospital inpatients who developed posttraumatic psychosis (as compared with brain-injured outpatient control subjects), this may sim- ply be an artifact of the selection process. Also, Sachdev et al.’s (2001) sample of patients with posttraumatic psycho- sis had more men than women, but this may simply be due to the greater prevalence of TBI in men. IQ/Cognition Although one recent study found no differences in IQ between brain-injured persons who went on to develop psychosis and those who did not (Fujii and Ahmed 2001), another recent study (Sachdev et al. 2001) found that the group that developed a schizophrenia-like psychosis had more neurological deficits than brain-injured control subjects, with lower IQ, significantly worse verbal and nonverbal memory, and greater impairments in language and frontal and parietal lobe functioning, consistent with a diffuse impairment in neuropsychological functioning. However, the authors acknowledge that it cannot be determined to what extent psychosis itself may have con- tributed to these deficits. Socioeconomic Status There are few data on the role of socioeconomic status in risk for posttraumatic psychosis. In one recent study, no differences in level of education attained was found between the group with psychosis secondary to TBI and a control group with TBI only (Fujii and Ahmed 2001). Substance Abuse There are few data on substance use or dependence as a risk factor for psychosis after TBI. In the newer case-control studies, there was more general previous substance use among those who developed posttraumatic psychosis (Fujii and Ahmed 2001) but no difference in use of psy- chosis-inducing substances such as lysergic acid diethyla- mide, amphetamines, and cocaine (Fujii and Ahmed 2001) and no difference in history of alcohol or drug dependence (Sachdev et al. 2001). Prior Neurological Disorder Fujii and Ahmed (2001) found that patients who went on to develop psychosis after a TBI had significantly more premorbid neurological pathology than did the brain- injured control subjects (80% vs. 40%; χ 2 =7.99; P <0.01), including prior brain injury (14/25), seizures (3/25), learning disability (3/25), birth complications (2/25), attention deficit hyperactivity disorder (1/25), and con- genital syphilis (1/25). This supports their hypothesis that psychosis may be more likely to follow TBI if the brain was already vulnerable before the injury. However, Sach- dev et al. (2001) did not find differences in perinatal or developmental abnormalities between the group that developed psychosis after TBI as compared with the brain-injured control subjects. Posttraumatic Epilepsy Delusions and hallucinations are known to be prevalent in temporal lobe epilepsy, which can result from brain injury (Flor-Henry 1969; Garyfallos et al. 1988; Lishman 1987; McKenna et al. 1985). A prospective study of patients with temporal lobe epilepsy found that 10% developed psychotic symptoms (Lindsay et al. 1979). A rigorous study in Iceland that involved clinical interviews found that 7% of epilepsy patients had psychotic symptoms (Gudmundsson 1966). Furthermore, patients with psy- 220 TEXTBOOK OF TRAUMATIC BRAIN INJURY chosis are 3–7 times more likely than the general popula- tion to have features of epilepsy, and interictal psychoses frequently resemble chronic schizophrenia. Hillbom (1960) found that the incidence of posttraumatic epilepsy in brain-injured Finnish veterans who developed psycho- sis was 57.5%, compared with only 31.8% in those with no psychiatric sequelae; however, the relationship between posttraumatic psychosis and epilepsy was not specific, because the incidence of posttraumatic epilepsy was 55.6% in the group of brain-injured veterans who had any significant psychiatric sequelae (psychotic and nonpsychotic). The more recent studies by Fujii and Ahmed (2001) and Sachdev et al. (2001) did not find a link between epi- lepsy and posttraumatic psychosis; in fact, Sachdev et al. found a trend toward less epilepsy in patients compared with control subjects. These findings appear paradoxical given that schizophrenia-like psychosis is 6–12 times more likely to occur in the context of epilepsy than in the general population (Sachdev 1998), and TBI is clearly known to be associated with the onset of seizures. It is reasonable to hypothesize that seizures could be a medi- ating phenomenon between TBI and psychosis, but the newer data do not support this theory. It may be that a longer time of follow-up after TBI might be needed to detect a relationship, because Davison and Bagley (1969) found that posttraumatic epilepsy was associated with de- layed onset of psychosis, as opposed to immediate onset of psychosis; the mean interval between onset of seizures and onset of psychosis was noted to be approximately 14 years. History of TBI in Patients With Schizophrenia A connection between TBI and subsequent psychosis is also supported by retrospective studies of premorbid brain injury in patients with schizophrenia, which reveal elevated rates of prior TBI compared with other groups. In a review of five studies published between 1932 and 1961, Davison and Bagley (1969) found the frequency of premorbid TBI in hospitalized patients with schizo- phrenia to range from 1% to 15%. This wide range of values likely derives from differences in definitions of brain injury and schizophrenia. Wilcox and Nasrallah (1987) reviewed the records for a history of TBI in 659 patients admitted to a large tertiary care center. Psychi- atric diagnoses were made according to research diag- nostic criteria, and TBI was defined as brain trauma occurring before age 10 years and resulting in either loss of consciousness for at least 1 hour or medical complica- tions (vomiting, confusion, visual changes). They found a premorbid history of TBI in 11% of patients with schizophrenia, compared with 4.9% of patients with mania, 1.5% of patients with depression, and 0.7% of surgical control subjects. Likewise, in a sample of Nige- rian patients diagnosed with research diagnostic criteria, patients with schizophrenia were found to have signifi- cantly more premorbid TBI than did patients with mania (Gureje et al. 1994). Malaspina et al. (2001) found a threefold greater rate of reported TBI for individuals with schizophrenia compared with their never mentally ill family members in a combined pedigree sample of families with bipolar disorder and schizophrenia, for a total of 1,832 members. (However, patients with schizo- phrenia were not significantly more likely to have incurred TBI than were patients with bipolar or depres- sive disorder.) In a replication, AbdelMalik et al. (2003) also found more childhood TBI among schizophrenia patients than in their unaffected siblings (OR=2.35; CI=1.03–5.36). Does Posttraumatic Psychosis Differ From Psychosis That Occurs Without Premorbid TBI? Atypical Versus Typical Symptoms One criterion listed in DSM-IV-TR for distinguishing psychosis secondary to a general medical condition from a primary psychotic disorder is the presence of atypical features such as visual and olfactory hallucina- tions (i.e., burning rubber or unpleasant smells). For example, there are case reports of Lilliputian hallucina- tions occurring in individuals with previous brain trauma (Cohen et al. 1994). Furthermore, there appears to be a link between right hemispheric injury and con- tent-specific misidentification delusions such as Capgras’ syndrome (loved ones replaced by identical- appearing impostors), Fregoli’s syndrome (persecutor able to change appearances and appear as different peo- ple), and reduplicative paramnesia (familiar place exists in two different places at the same time) (reviewed in Edelstyn and Oyebode 1999; Forstl et al. 1991; McKenna et al. 1985) However, only between 25% and 40% of cases of Capgras’ syndrome are associated with neuro- logical disorders, so such atypical symptoms are not pathognomonic for psychosis due to a general medical condition. Additionally, posttraumatic psychosis fre- quently occurs without these atypical symptoms. For example, in a study of 45 individuals with schizophrenia- like psychosis after TBI, none of the sample demon- Psychotic Disorders 221 strated misidentification syndromes, only 15% had reli- gious delusions, 20% had visual hallucinations, and 4% had tactile hallucinations (Sachdev et al. 2001). In con- trast, 55% of these patients with posttraumatic schizo- phrenia-like psychosis had persecutory delusions and 84% had auditory hallucinations, which are common symptoms in schizophrenia. The low rates of atypical psychotic symptoms and high rates of typical symptoms in the Sachdev et al. (2001) study may be related to the study design, because individuals had to meet DSM-IV- TR Criteria A, B, C, and E for schizophrenia or schizo- phreniform disorder to be included. A more inclusive sample of any posttraumatic psychosis might demon- strate more atypical and fewer typical psychotic symp- toms. However, others have also reported that paranoia and delusions are common symptoms in post-TBI psy- chosis (Cutting 1987). In contrast to the overlap of positive symptoms of psy- chosis, only 22% of Sachdev et al.’s (2001) sample dis- played negative symptoms (such as flattening of affect, avolition, or asociality), and only 4% had derailment or thought disorder. This is consistent with previous reports of relative absence of formal thought disorder and of blunting of affect in schizophrenia after TBI (McKenna 1994). However, the finding of low rates of negative symptoms is not consistent with the study by Thomsen (1984), which found that patients who developed psycho- ses after severe blunt brain trauma often developed deficit types of symptoms, including anhedonia, aspontaneity, and loss of social contact, probably related to the high rate of frontal injuries. The course of psychotic illness among the brain-injured individuals with psychosis in the Sachdev et al. (2001) study was similar to that of schizophrenia not associ- ated with TBI, because the patients had prodromal symptoms such as scholastic or work deterioration and social withdrawal, with a gradual onset of psychotic symptoms at a similar age accompanied frequently by depression (50%) and a subsequent subacute or chronic course. Cognition As with positive and negative symptoms, there is no clear consensus as to whether posttraumatic psychosis can be differentiated from primary psychotic disorders by the extent of cognitive impairment. In a Nigerian sample of patients with schizophrenia, those with a his- tory of childhood brain trauma that required hospital- ization had poorer scholastic performance as children (Gureje et al. 1994). They were also found to have mixed laterality as adults, possibly due to left hemi- spheric damage. However, we have found (Corcoran et al. 2000) that among patients with schizophrenia, those with a history of TBI actually had better cognition than those who did not. Family History/Genetic Vulnerability An early study suggested that brain trauma could con- tribute to schizophrenia either 1) directly or 2) through an interaction with latent vulnerability, and that these two pathways yielded different symptom patterns (Sha- piro 1939). Shapiro (1939) evaluated 2,000 cases of dementia praecox (schizophrenia) in residents of a large public hospital and found that “a large number . . . showed some relationship to a severe head injury.” To establish a sample in which there was less doubt that the brain injury and psychosis were linked, he selected 21 cases in which the schizophrenia-like psychosis quickly ensued after the brain injury, beginning within a few hours to 3 months afterward. Ten of the 21 patients had no grossly obvious signs suggestive of the sequelae of the trauma; all 10 of these patients demonstrated a predispo- sition to schizophrenia such as positive family history or “introverted” premorbid personality. Shapiro concluded that in these 10 patients, the brain trauma acted as a pre- cipitating factor. The other 11 patients showed symp- toms not only typical of schizophrenia but other “neuro- logical” features as well, such as headache, seizures, confusion, dizziness, disorientation, and memory impair- ment. In this group, only 2 of the 11 showed “hereditary tainting,” and 7 of the 11 had “well-integrated” premor- bid personalities. Shapiro concluded that in this group, brain trauma not only precipitated but directly contrib- uted to the etiology of the psychosis. Other studies have suggested that TBI can contrib- ute to schizophrenia risk, because among schizophrenia patients, those without premorbid TBI have more ge- netic vulnerability for psychotic disorders than do those with prior TBI, who have no greater rates of family members with psychosis than do the general population (Davison and Bagley 1969). In a reexamination of a data- base of 722 probands with schizophrenia (originally studied by Rudin), the diagnosis of schizophrenia was confirmed in a subsample of 660, and the prevalence of schizophrenia in the parents and siblings of these 660 probands was examined (Kendler and Zerbin-Rudin 1996). It was found that the risk for schizophrenia was particularly low in siblings of probands whose onset of illness occurred within a year of major brain trauma. Malaspina et al. (2001) found that TBI may interact with schizophrenia genetic vulnerability to increase the risk for schizophrenia. 222 TEXTBOOK OF TRAUMATIC BRAIN INJURY What Are Common Cognitive Features of TBI and Schizophrenia? The presence of similar features in TBI and schizophre- nia may shed light on the pathophysiological mecha- nisms by which these phenomena may be associated. Key similarities between TBI and schizophrenia include deficits in insight, executive function, and memory, which indicate pathology in similar neuroanatomical sites, such as, respectively, the orbitofrontal regions, dorsolateral prefrontal cortex, and hippocampi. Com- mon deficits in sensory gating may implicate abnormal connectivity between various parts of the brain in both conditions. Poor Insight Up to one-half of individuals with moderate to severe TBI have reduced awareness of their deficits (Flashman et al. 1998; see Chapter 19, Awareness of Deficits). Poor insight is highly prevalent in schizophrenia patients and is characterized by deficits in awareness of having a men- tal disorder, of response to medication, of the social con- sequences of the mental disorder, and of specific symp- toms of the illness (Amador et al. 1994; Pini et al. 2001). Poor insight complicates compliance with treatment recommendations in both those with brain injury and those with psychotic disorders. Neuropsychological Function Cognitive deficits are common in both brain-injured individuals and those with schizophrenia. Impairments in executive functions occur frequently in both groups, such as planning and problem solving needed for activ- ities such as balancing bank accounts, writing letters, planning one’s week, and driving or taking public transportation (Mazaux et al. 1997). Formal neurocog- nitive tests of executive function include the Trail Mak- ing Test B, Wisconsin Card Sorting Test, and Tower of Hanoi. Poor performance on these tests is a common finding both in individuals with a TBI (Brooks et al. 1999; Callahan and Hinkebein 1999; Leon-Carrion et al. 1998; Wiegner and Donders 1999) and in individu- als with schizophrenia (reviewed in Johnson-Selfridge and Zalewski 2001). Individuals with both schizophre- nia and brain injury also show deficits in explicit mem- ory, which is the deliberate recall of facts such as dates and phone numbers, as well as decrements in volume of the hippocampus, the part of the brain thought to be responsible for explicit memory. In both groups of patients, the extent of memory deficit is associated with the degree of volume reduction of the hippocampus (Gur et al. 2000; Tate and Bigler 2000). Neuroanatomical Effects of TBI and Implications for Psychosis Pathophysiology Perhaps accounting for the overlap in cognitive deficits seen in both groups, there is significant overlap between the brain regions implicated in schizophrenia and those regions that are vulnerable to TBI, including the frontal and temporal cortices and the hippocampus. Primary Sites of Lesion Brain injury frequently results in damage to the frontal and temporal cortices. Similar regions are often involved in individuals who develop psychosis from other neuro- logical conditions such as metachromatic leukodystrophy and cerebrovascular disease (Buckley et al. 1993; Hyde et al. 1992; Levine and Grek 1984; Miller et al. 1991; Rabins et al. 1991; Richardson 1992). In epilepsy, visual halluci- nations have been found to result from seizure foci in the temporal lobes or orbitofrontal regions (Fornazzari et al. 1992) and delusions of passivity (“forces are acting upon me,” “I am being controlled”) have been linked to left temporal lobe seizure foci (Perez and Trimble 1980; Trimble and Thompson 1981). Of interest, in early experiments of stimulation of the brains of awake patients undergoing neurosurgery, stimulation of the temporal lobes elicited auditory hallucinations (Mullan and Pen- field 1959). Abnormalities in the prefrontal cortex are common in schizophrenia, and it has been hypothesized that the attendant working memory deficits (holding information online while attending to other tasks) may be the key pathophysiological feature of schizophrenia. Secondary Sites of Lesion Brain injury also results in damage to regions far from the primary site of impact (diaschisis) (Joashi et al. 1999). Ani- mal studies of TBI, including weight-drop and fluid per- cussion models, show that the hippocampus is particularly vulnerable to TBI, even injuries that have a primary impact far from the hippocampus (Bramlett et al. 1997; Chen et al. 1996; Colicos et al. 1996; Lowenstein et al. 1992; Qian et al. 1996; Tang et al. 1997b; Yamaki et al. 1998). Furthermore, hippocampal injury in animals leads to memory impairments (Chen et al. 1996; Tang et al. Psychotic Disorders 223 1997a). Of note, the cell loss in the hippocampus is pro- gressive longer than 1 year after TBI in rats, suggesting a possible explanation for what is observed in humans: ongoing changes in the brain months to years after the initial injury (i.e., a chronically progressive degenerative process initiated by brain trauma) (Smith et al. 1997a). This special vulnerability of the hippocampus to trauma may be due to axon stretching and diffuse axonal injury, which are common features of brain trauma in an- imals and humans. When diffuse axonal injury was repli- cated in pigs through nonimpact inertial loading, there was widespread multifocal injury observed of axons and neurons, especially in regions of the hippocampus (Smith et al. 1997b). In nonhuman primates with acceleration- induced experimental brain injury, 59% developed hip- pocampal lesions: 46% of animals with mild injury (brief unconsciousness and no residual neurologic deficit) and 94% of animals with severe injuries. Cell death in the hip- pocampus occurred without a drop in cerebral perfusion pressure or increase in intracranial pressure and did not seem to be a consequence of low oxygen, because other regions of the brain vulnerable to hypoxia did not have cell death (Kotapka et al. 1991). Traumatic injury to the hippocampus also occurs in humans in the absence of el- evated intracranial pressure (Kotapka et al. 1994). Abnormalities in hippocampal structure and function are common in schizophrenia. A meta-analysis of 18 stud- ies showed a bilateral reduction of volume in the hippo- campus in schizophrenia of 4% (Nelson et al. 1998). Magnetic resonance spectroscopy studies suggest that neuronal integrity is compromised in the hippocampus in schizophrenia, because low N-acetylaspartate has been found across several studies (reviewed in Poland et al. 1999 and Soares and Innis 1999). Silbersweig et al. (1995) found increased blood flow in the hippocampus during hallucinations. Postmortem studies provide evidence that there is synaptic and, hence, circuitry abnormality in both the hippocampus and the prefrontal cortex (Harrison 1999). Intriguingly, cognitive and magnetic resonance imaging volumetric assessments of twins discordant for schizophrenia suggest that hippocampal abnormality is more prevalent in the affected twin, suggesting nonge- netic influences operating on the hippocampus in schizo- phrenia (Baare et al. 2001; Cannon et al. 2000; Suddath et al. 1990). Disturbances in connectivity among different regions of the brain are a common result of TBI and have been hypothesized to play a role in the genesis of some symp- toms of schizophrenia. For example, Frith (1996) sug- gested hallucinations result from disruption in connectiv- ity among parts of the brain responsible for intentional speech and observation/interpretation of speech, so that auditory sensory phenomena are misattributed to exter- nal sources. Furthermore, TBI can impair the ability to filter incoming sensory information; deficits in the gat- ing/filtering of sensory information are also characteristic of schizophrenia. It has been hypothesized that these ab- normalities result from disruptions in connections be- tween different parts of the brain, and that the inability to filter out stimuli can lead to sensory “flooding” by irrele- vant information. Populations Who Are Vulnerable to Posttraumatic Psychosis Homeless Individuals Homeless people have high rates of schizophrenia-like psychosis and TBI history (Silver and Felix 1999). Studies have shown that homeless persons have an elevated prev- alence of schizophrenia that ranges between 13.7% (Koe- gel et al. 1988) and 25% (Susser et al. 1989). More than 40% of homeless individuals with a schizophrenia-like psychosis who were treated at a university hospital in New York had a history of premorbid TBI (Silver and McKinnon 1993). Death Row Prisoners An interesting study of 15 death row inmates showed that all 15 had histories of severe brain injury and 9 had recur- rent psychoses (with hallucinations, delusions, thought dis- order, and bizarre behavior) that antedated incarceration (Lewis et al. 1986). Remarkably, these subjects were not selected for clinical evaluation because of any evident psy- chopathology but rather were chosen for neuropsycholog- ical testing in the hope of appealing for clemency when their executions were imminent. That is, these were indi- viduals who had not been identified as mentally ill but who were at the final stages of their appeals process. All had repetitive episodes of brain trauma beginning in childhood that were quite dramatic—severe physical abuse, falling from heights, being hit by and run over by cars, being hit with baseball bats. The episodes of brain trauma were cor- roborated by scars, indentations of the cranium, hospital records, and CT scans. They had comprehensive evalua- tions by a board-certified psychiatrist lasting from 4 to 16 hours that involved detailed birth, development, neurolog- ical, psychiatric, medical, educational, family, and social histories; interviews of family members; physical examina- tions; CT scans; and electroencephalography. The inmates largely tried to conceal their psychotic symptoms. Of note, all but one had a normal IQ. 224 TEXTBOOK OF TRAUMATIC BRAIN INJURY Children and Teens The National Institutes of Health Consensus Develop- ment Panel on Rehabilitation of Persons With Traumatic Brain Injury (Consensus conference 1999) reports the highest incidence of brain trauma is among individuals 15–24 years old (and the elderly), with another peak in children younger than 5 years. Motor vehicle accidents are the major cause of TBIs in the 15- to 24-year-old group, and alcohol is frequently involved. Sports injuries and violence also are a major cause of brain injury in teens. Child abuse and assault is also a significant cause of TBI in children. Of note, reported rates of prior child abuse are 20/38, or 52%, of patients with first-episode psychosis (Greenfield et al. 1994) and 27/61, or 44%, of patients with chronic psychosis (Goff et al. 1991). Evaluation of Posttraumatic Psychosis A thorough assessment of the patient with posttraumatic psychosis is an essential prerequisite to the prescription of any treatment (Arciniegas et al. 2000). A comprehensive evaluation must include detailed histories of birth, devel- opment, neurological features, psychiatric symptoms, medical status, education, substance use, social function- ing, and any family illnesses, as well as physical and neu- rological examinations, detailed mental status examina- tion, neuropsychological testing using a standardized battery, structural imaging (CT or magnetic resonance imaging), and electroencephalography. Premorbid his- tory and current medication treatment are important because they can influence neuropsychiatric symptoms (Arciniegas et al. 2000). Family members and other cor- roborating sources should be included in the examination because individuals may not recall details of brain injury if it occurred either when they were children or when they were intoxicated, and the neuropsychological corre- lates of both psychosis and TBI can interfere with the ability to recall one’s history in detail (McAllister 1998). Posttraumatic Amnesia In the initial period after injury, during the period of PTA, numerous features of delirium are likely to occur (see Chapter 9, Delirium and Posttraumatic Amnesia), including restlessness, fluctuating level of consciousness, agitation, combativeness, emotional lability, emotional withdrawal or excessive dependency, confusion, distracti- bility, disorientation, and amnesia (Trzepacz 1994). Hal- lucinations and delusions may also occur during this period, although delusions are seldom well organized (Goethe and Levin 1984; McAllister and Ferrell 2002; Trzepacz 1994). Expressive and receptive speech and lan- guage disturbances, including perseveration, are fre- quently present during this period and can produce a clin- ical picture similar to the disorder of thought and language found in schizophrenia (Goethe and Levin 1984). Many of these symptoms are likely to improve as the period of PTA improves. Posttraumatic Epilepsy Psychotic syndromes associated with posttraumatic epi- lepsy occur in the peri-ictal period (either during seizures or in the immediate postictal period) or interictally, in which case the psychotic symptoms are more commonly chronic rather than episodic (McAllister and Ferrell 2002; Trimble 1991). The most common of these entities is the postictal acute confusional state characterized by general- ized confusion, fluctuating sensorium, agitation, halluci- nations, and delusions, which is similar to the posttrau- matic delirium described in the preceding section. This condition generally resolves within a few hours after the seizure, although it may rarely persist for several days. It is important to detect whether the patient has a seizure disorder, because this can be treated with anticonvulsants and also because so many psychiatric medications can lower the seizure threshold. Mood Disorders Mood disorders are a common occurrence after TBI, and both depression and mania can present with psychotic symptoms. Manic syndromes with associated psychosis and schizoaffective syndromes after TBI have been described largely in single case reports or small series. Shukla et al. (1987), for example, reported on 20 patients with manic or schizoaffective symptoms and a history of TBI. In this series, psychotic symptoms occurred in a high percentage of patients. Grandiosity occurred in 90%, pressured speech in 80%, and flight of ideas in 75%. No one in this series had a positive family history for bipolar disorder, indicating that genetic loading is not a necessary prerequisite for development of mania after brain injury. Psychotic symptoms are prominent in many of the cases of mania subsequent to TBI reported in the literature (Bracken 1987; Clark and Davison 1987; Nizamie et al. 1988; Pope et al. 1988; Reiss et al. 1987). Depression is more common than mania after TBI and can also be associated with psychotic symptoms in approximately 25% of individuals (Hibbard et al. 1998; McAllister and Ferrell 2002). Obviously, it is important to recognize mood disorders as the cause of psychotic symp- Psychotic Disorders 225 toms, because the treatment follows logically from this diagnosis. Treatment of Posttraumatic Psychosis Any existing delirium, seizure disorder, mood disorder, or substance abuse or dependence must be diagnosed and attended to in the treatment of posttraumatic psychosis. If these disorders are not present, if psychotic symptoms are life-threatening, or if psychotic symptoms persist beyond the treatment of these disorders, then an antipsychotic medication may be warranted. Care should be taken in administering neuroleptics, as animal studies suggest that dopamine antagonists (antipsychotic medications) can impede recovery after brain injury (Feeney et al. 1982). Problems with motor function, gait, arousal, and speed of information processing are common in brain-injured patients and may be exacerbated by the sedation, psycho- motor slowing, parkinsonism, and anticholinergic side effects of neuroleptics. Of note, there are no controlled studies of treatments for psychosis in patients with pre- morbid TBI. Information comes from case reports and extrapolation from studies in other populations of patients with brain damage. Given these caveats, most cli- nicians advise that neuroleptics should be used specifically for psychotic symptoms and not for agitation only. Medication dosing should be “low and slow.” Many experts suggest starting with one-third to one-half of the usual dose (McAllister 1998). The clinician must be wary of medications with significant sedative and anticholiner- gic properties. Therefore, among typical neuroleptics, high-potency antipsychotic medications such as haloperi- dol (Haldol) may produce fewer of these side effects than low-potency antipsychotics such as chlorpromazine (Thorazine). However, it should be noted that TBI may also make patients more vulnerable to developing tardive dyskinesia (Kane and Smith 1982). Atypical antipsychotic drugs have emerged as first- line drugs for treatment of psychotic disorders. These drugs offer two main advantages over conventional neu- roleptic drugs. They have greater efficacy, especially in decreasing negative as well as positive symptoms of schizophrenia and in decreasing agitation and aggression. The latter effect can be of particular benefit in some indi- viduals with TBI. Most important, the atypical antipsy- chotics carry significantly less risk of causing extrapyra- midal symptoms (EPSs) and tardive dyskinesia. Like all drugs with antipsychotic activity, the atypicals have some blocking effect on dopamine-2 receptors but proportion- ally less so than conventional drugs. The atypical class also shows a preference for limbic dopamine-2 receptors with minimal nigrostriatal effects, and thus less risk of EPSs. Clozapine is a candidate for the treatment of posttrau- matic psychosis in that it yields a low incidence of EPSs and tardive dyskinesia. Case reports of clozapine suggest efficacy in patients with posttraumatic psychosis. For ex- ample, 400 mg of clozapine daily was effective for a 34- year-old man who had a 10-year history of refractory and persistent voices and delusions after a brain injury at age 12 years (Burke et al. 1999). However, a less clear picture was observed in an open trial of clozapine in a series of nine brain-injured patients with either refractory psy- chotic symptoms or treatment-resistant outbursts of rage and aggression (Michals et al. 1993). In this series, one- third of patients had, respectively, marked improvement, mild improvement, and indeterminate improvement. However, seizures occurred in two of the nine patients, including new onset of seizures in one patient who was taking 600 mg/day of clozapine, along with pimozide and amoxapine. The other patient had a preexisting seizure disorder and developed a recurrence while taking low doses of clozapine (75–100 mg/day) despite also taking an anticonvulsant (valproate, 4,000 mg/day) and a benzodi- azepine (lorazepam, 3 mg/day.) These data suggest that clozapine should be given pri- marily to individuals with posttraumatic psychosis with- out a history of seizures, and that prophylactic anticon- vulsants such as valproate may be indicated to prevent the onset of new seizures. Clozapine can also cause sedation and dizziness, for which brain-injured patients may have greater vulnerability. Additionally, there are risks of agranulocytosis (minimized with weekly blood draws), ta- chycardia, orthostatic hypotension, hypersalivation, and weight gain. Because of this side-effect profile, clozapine is not usually the first of the atypical antipsychotics to try. We suggest trying at least two of the other atypical anti- psychotic drugs before beginning a clozapine trial. Among the other atypical antipsychotics, none shows clearly superior efficacy. A particular patient’s history of previous response, minimizing certain side effects, and the clinician’s familiarity with one drug or another all af- fect choice of drug. In some instances, one might wish to use a side effect such as sedation or tendency to cause weight gain to advantage. One should strive to make one change at a time to prevent confusion about the cause of subsequent clinical changes. Ineffective drugs should be discontinued. When adding a drug to the regimen, con- sider stopping the current drug to avoid polypharmacy. A variety of case reports and small case series suggest that most of the atypical antipsychotics, including olanza- pine, risperidone, and quetiapine, can be used to effec- tively treat psychosis resulting from TBI, although there This page intentionally left blank [...]... recruited from brain injury organization newsletters, PTSD was the most common anxiety disorder reported (Hibbard et al TEXTBOOK OF TRAUMATIC BRAIN INJURY 1998) Of the 17% of the subjects who reported PTSD developing after injury, 41 % of them had experienced resolution of their symptoms by the time of the interview Subjects in this study were approximately 7.6 years postinjury Approximately one-half of the... may be perceived as childish One component of this type of childish behavior relates to the Eriksonian stage (Table 13 4) that is present at the highest risk period for the occurrence of TBI (15 to 24 years old) At that age, the stage of identity versus dif- TEXTBOOK OF TRAUMATIC BRAIN INJURY TABLE 13–3 Pragmatic language dysfunction after traumatic brain injury Decreased intelligibility Choppy rhythm... type of occurrence) Acute 6 months 33.7 (aggression) 40 (restless) 19.0 Rehabilitation — 42 .0 55 1 year 71 67.0 Severe 42 5 years 64 64. 0 Oddy et al 1985 Severe 44 7 years 43 31.0 Thomsen 19 84 Severe 40 2–5 years 38 — Thomsen 19 84 Severe — 10–15 years 48 — Van Zomeren and Van Den Berg 1985 Severe 57 2 years 39 — Levin et al 1979 Severe 27 1 year 37 — McMillan and Glucksman 1987b Moderate 24 — 64 — Schoenhuber... patients with a mean of 8 years between TBI and structured clinical interview Patients were recruited by advertisements in brain injury newsletters in New York Severity of brain injury ranged widely, with 40 % of patients having severe TBI by self-report The authors found that Posttraumatic Stress Disorder and Other Anxiety Disorders TABLE 12–1 Rates of anxiety disorders after traumatic brain injury, from case... primary diagnosis of TBI who met the definition of agitation (Wolf et al 1996) In a series of 67 patients admitted with mild to moderate TBI and rated prospectively, restlessness occurred in 40 % and agitation occurred in 19% (van der Naalt et al 2000) Studies of mild TBI have evalu- 259 260 TEXTBOOK OF TRAUMATIC BRAIN INJURY ated individuals for much briefer periods of time: 1-year estimates of irritability,... to the brain injury directly or to the accumulation of severe life experiences that immediately follow the brain injury or to the combination of both events Because the pathophysiology of anxiety disorders remains unknown, only inferences can be made about the contribution of the brain injury to the development of posttraumatic anxiety The temporal association of the TBI with the development of anxiety... anatomic areas of the brain are important in the production (or lack of suppression) of “irritative aggression,” that is, feelings of irritability with occasional explosions Table 14 4 summarizes the roles of key regions of the brain in mediating aggression Hypothalamus Many areas of the brain are involved in the production and mediation of aggressive behavior, and lesions at different levels of neuronal... DC, American Psychiatric Association, 2000 Used with permission Pathophysiology of Aggression Neuroanatomy of Aggression Many areas of the brain are involved in the production and mediation of aggressive behavior, and lesions at dif- 262 TABLE 14 4 TEXTBOOK OF TRAUMATIC BRAIN INJURY Neuropathology of aggression Locus Activity Hypothalamus Orchestrates neuroendocrine response via sympathetic arousal, monitors... lability, impairment of impulse control, and suspiciousness (Garyfallos et al 1988) Neocortex The most recent region of the brain to evolve, the neocortex, coordinates timing and observation of social cues, often before the expression of associated emotions Because of the location of prominent bony protuberances in the base of the skull, this area of the brain is highly vulnerable to traumatic injury Lesions... following section focus on the topic of PTSD after TBI with amnesia ≤25 Social phobia Posttraumatic stress disorder Not known 0 42 80% of patients met the criteria for a DSM-IV-TR mental disorder on the basis of patients’ reports after their brain injury The validity of these retrospective diagnoses, however, may be questioned because of the long duration between injury and interview Moreover, patients’ . views of the author(s) and are not to be construed as of cial or as reflecting the views of the Department of the Army or the Department of the Defense. 232 TEXTBOOK OF TRAUMATIC BRAIN INJURY tiple. rates could represent a selection bias of patients pre- 2 34 TEXTBOOK OF TRAUMATIC BRAIN INJURY senting for brain injury rehabilitation. Anxiety disorder pa- tients also had greater medical and. reports of clozapine suggest efficacy in patients with posttraumatic psychosis. For ex- ample, 40 0 mg of clozapine daily was effective for a 3 4- year-old man who had a 10-year history of refractory