ANPA Abstracts, Poster Sessions I and II

POSTER SESSION I—SUNDAY, MARCH 10

General Clinical Neuropsychiatry

P1. Akinetic mutism following anterior cerebral artery occlusionC. Alan Anderson, David B. Arciniegas, Daniel C. Huddle, Maureen Leehey (University of Colorado Health Sciences Center, Denver, CO)

Background: Akinetic mutism is a disturbance of motivation typically associated with bilateral mesial frontal lesions. Although the syndrome has been described with unilateral lesions, the pathophysiology is unknown. Case Report: A 61-year-old man presented to the emergency department with acute akinetic mutism. On examination, he was alert but inattentive. Spontaneous movement and speech were absent. With repeated prompting, he followed single-step commands only intermittently. Diagnostic neuroimaging studies were performed. An arteriogram demonstrated occlusion of the left anterior cerebral artery. MRI including diffusion-weighted sequences demonstrated an infarction in the distribution of the left anterior cerebral artery involving the cingulate gyrus and the left half of the corpus callosum. Conclusions: Although typically a consequence of bilateral anterior cingulate/callosal injury, the syndrome of akinetic mutism may also result from a lesion of the cingulate gyrus lesion in one hemisphere when extension into the body of the corpus callosum effectively disconnects the contralateral cingulate gyrus. This case is the first in vivo description of acute akinetic mutism produced by such a lesion.

P2. Is self-injurious behavior in patients with severe developmental disabilities an addiction?L. Jarrett Barnhill (University of North Carolina School of Medicine, Chapel Hill, NC)

Background: Self-injurious behavior (SIB) is a heterogeneous behavior that spans the range of psychiatric and neurological disorders. SIB may occur in 15% to 20% of patients with autism and severe developmental disabilities. Some milder forms of SIB are related to high rates of repetitive behaviors. Severe SIB appears driven and compulsive, and it continues in spite of severe tissue damage. Both subtypes share characteristics with addiction behaviors. Methods: Case studies will be provided from the Developmental Neuropharmacology Clinic and multiple residential programs for patients with severe developmental disabilities. The typology of SIB will be compared with addiction behaviors. Results: Milder forms of SIB resemble stereotypies and compulsive spectrum disorders. They appear to be preceded by an urge to perform the repetitive behavior, and they appear intrinsically reinforcing. Severe SIB can be compulsive and relentless, and even severe pain fails to interrupt the behavior. For many patients, pain or specific environmental or affective states trigger the SIB, but once initiated, it continues in an out-of-control fashion. Conclusions: SIB has features consistent with drug addiction. Milder forms of SIB (high frequency/low intensity) resemble the early phases of drug abuse. More severe SIB shares features of progressive addiction: craving, loss of control, tolerance to intrinsic reinforcement, and sensitization. The discussion will focus on the similarities between SIB and sensitization from chronic psychostimulant abuse and the common ground between neurobiological substrates and treatment strategies.

P3. Vagal nerve stimulation for medically refractory epilepsy in persons with severe developmental disordersL. Jarrett Barnhill, Sue Lannon, Bradley Vaughn (University of North Carolina School of Medicine, Chapel Hill, NC)

Background: Medically refractory epilepsy, developmental brain abnormalities, and neurobehavioral disorders are commonplace in individuals with severe developmental disabilities. Vagal nerve stimulation (VNS) is a promising strategy for refractory epilepsy and mood disorders in patients with developmental disabilities. Methods: This study included 31 autistic and intellectually disabled patients who received a VNS. We compared changes in seizure frequency and neuropsychiatric status at baseline with one-year post-VNS follow-up. Results: Of the 31 patients who received a VNS, 15 had slight or moderate reduction in seizure frequency. Few patients were able to discontinue or decrease doses of their antiepileptic drugs. Neurobehavioral deterioration occurred in 6/12 VNS nonrepsonders and 5/6 patients with comorbid autism. Negative behavioral outcomes included increased irritability/moodiness, aggression, and social withdrawal. Conclusions: VNS is an exciting technique for patients with medically refractory epilepsy. In our heterogeneous sample, VNS appeared partially effective in 15/31 patients. Improvements in quality-of-life measures occurred with reductions in seizure frequency and included increased alertness as well as reduced frequency. Our results suggest that clinical improvement with VNS is less predictable in patients with severe developmental disabilities. Patients with comorbid autism appear at greater risk for neurobehavioral complications.

P4. Levodopa treatment for tics: preliminary reportKevin I. Black, Johanna M. Hartlein, Bradley L. Schlaggar, Jonathan W. Mink (Washington University School of Medicine, St. Louis, MO; University of Rochester, Rochester, NY)

Background: Because of the clear response of tics to neuroleptics, it was initially thought that tics reflected dopaminergic hyperactivity. However, tics do not diminish if comorbid Parkinson's disease develops, and we recently reported that a single-blind acute dose of levodopa 150 mg actually diminished tic severity by an average of 40% in 6 adults with TS. This is an interim report from a randomized controlled trial. Methods: 17 volunteers with Tourette syndrome or chronic motor tic disorder, with no recent neuroleptic treatment, have enrolled thus far in a double-blind treatment study. Levodopa/carbidopa 100/25 mg, or matching placebo, is given in doses of 0.5–2.0 tabs po tid (0.5–1.0 in prepubescent children). Results: Seven patients are ages 8–17 and 10 are adults. Complete data through the first 8 weeks of treatment are available from 13 subjects, except that we remain blind to drug assignment. Three subjects were withdrawn because of noncompliance (one after week 8), and 2 dropped out due to worsened impulsivity or anger. Possible side effects in the other 12 subjects have been mild and include nausea (1), transient hypertension (1), migraine (in a subject being treated for migraines), and mild worsening of tics (1). Clinical global impression of tic change by a blinded physician rater ranged from 2 in one subject to 5 in two subjects, with a mean of 3.7 (2=moderate improvement, 3=minimal improvement, 4=unchanged, 5=minimal worsening). Pill counts were consistent with excellent compliance in 10 of 13 subjects. Mean self-rated tic severity improved slightly from 17.8 at baseline to 13.8 at week 8. Conclusions: Consistent with our earlier report, substantial doses of levodopa do not cause significant worsening of tics. This finding does not support a hypothesis that tics result from dopaminergic hyperactivity. Further analysis awaits eventual unblinding.Support from Tourette Syndrome Association and NINDS.

P5. Aggressive behavior after a recent stroke: initial findings and six-month follow-upKeen-Loong Chan, David Moser, Stephan Arndt, Robert G. Robinson (University of Iowa, Iowa City, IA; Tan Tock Seng Hospital, Singapore)

Background: Aggression is often seen in neuropsychiatric conditions. We attempted to determine the correlates of aggression after stroke and to follow up on the progression over 6 months. Methods: 92 patients with recent stroke were assessed for aggression on the basis of symptoms elicited by the Present State Examination (PSE). Aggressive patients were compared with nonaggressive patients on total PSE scores, Ham-D, Ham-A, MMSE, Johns Hopkins Functioning Inventory, psychiatric symptoms, frontal-executive functions, and CT scan findings. Results: 25% (23/92) of patients were aggressive. Aggressive patients had higher total PSE, Ham-D, and Ham-A scores, lower MMSE scores, and lesions that were more proximal to the frontal pole of the brain than nonaggressive patients. There were no differences in tests of frontal-executive function or measures of brain atrophy. Stepwise regression showed that Ham-A and lesion proximity to the frontal pole were significant in predicting aggression. By 6 months, most (22/23) of the aggressive patients had improved in their aggression, irrespective of treatment with antidepressant or placebo. Conclusions: In the majority of cases, early-onset aggression after a recent stroke appears to improve irrespective of treatment with antidepressant or placebo. Because anxiety significantly predicted aggression, more research is needed to determine if treatment with anxiolytics would reduce the incidence and severity of aggressive behavior after stroke.Support from NIMH Grants MH40355, MH52879, and MH53592.

P6. Evolution of primary progressive aphasia due to Alzheimer's diseaseDavid G. Clark, Mario F. Mendez* (University of California at Los Angeles, Los Angeles, CA)

Background: In primary progressive aphasia (PPA), there are isolated and progressive language deficits for at least two years prior to possible dementia (Neary et al, Neurology 1998; 51:1546–1554). The two major forms of PPA are a nonfluent aphasia and a fluent “semantic dementia” with loss of word meaning. The most common pathology in PPA is frontotemporal dementia, but approximately 20% have Alzheimer's disease (AD). We describe an AD patient who evolved from semantic dementia to nonfluent aphasia to reiterative speech. Case Report: A 70-year-old man presented with language deficits consistent with semantic dementia. He had a fluent and grammatical speech, paraphrasias, and impairments in naming and word comprehension. Over the next few years he developed hesitant, broken, telegraphic, and effortful speech. Near the end of his life, he had dysarthria, palilalia, and profound stuttering such that all he could utter were the first syllables of words. On autopsy, he had AD with neurofibrillary tangles and senile plaques involving the language areas. Conclusions: In AD, the syndrome of PPA can evolve through a spectrum of speech and language difficulties when pathology involves the perisylvian language areas. We discuss the diagnosis of PPA and management with speech therapy and dopaminergic drugs.

P7. Acute frontotemporal personality from hypoxic encephalopathyNima A. Fahimian, Margaret Swanberg, Mario F. Mendez* (University of California at Los Angeles, Los Angeles, CA)

Background: Severe disinhibition and the symptoms of the Klüver-Bucy syndrome are manifestations of frontotemporal dementia. Other conditions may cause these behavioral changes. We report an unusual case of a man with severe frontotemporal behaviors due to hypoxic encephalopathy. Case Report: An 82-year-old man was in his usual state of health when he awoke with an acute personality change associated with a probable cardiac arrhythmia. He went from being an introvert to an extravert and showed decreased impulse control with verbal outbursts. He became hypersexual and would demarcate his living room with feces. He exhibited hyperoral behaviors, hypermetamorphosis, labile mood and exophoria, and a tendency to explore by touching. The patient underwent neurobehavioral and neurological examinations and imaging studies. Examination disclosed evidence of frontal-executive deficits, attention difficulty, and frontal grasp reflexes. MRI revealed generalized atrophy, and PET showed hypometabolism more prominent in bilateral superior temporal gyrus followed by frontal regions. Conclusions: This patient illustrates the dramatic development of frontotemporal personality including the spectrum of frontal changes and anterior temporal Klüver-Bucy symptoms. We discuss the differential diagnosis of frontally predominant behavior changes and the Klüver-Bucy syndrome, in this case most consistent with postanoxic dystrophy.

P8. Critical periods of suicide risk in Huntington's diseaseKarin C. Ferneyhough, Laura M. Stierman, Beth M. Turner, Jane S. Paulsen,* and the Huntington Study Group (University of Iowa, Iowa City, IA)

Background: It is well established that suicide risk in patients with Huntington's disease is greater than that in the general population. No known large-scale study has examined differences in suicide risk across stages of HD. Methods: Suicidal ideation was examined in 4,342 individuals in the Huntington Study Group database. All subjects were grouped according to the standardized neurological exam from 0 (normal exam) to 3 (unquestionable HD). Patients with an unequivocal diagnosis of HD were further divided into stage of disease (stage 1 to stage 5) by using Shoulson and Fahn's total functional capacity measure. Results: The proportion of persons with suicidal ideation varied by diagnosis and HD stage, suggesting two primary critical periods during which suicidal ideation increased dramatically. The frequency of suicidal ideation doubled from 10.4% in at-risk persons with a normal neurological exam to 20.5% in at-risk persons with soft neurological signs (χ²=40.99, P<0.001). In persons with a diagnosis of HD, 15.7% had suicidal ideation in Stage 1, whereas 20.9% had suicidal ideation in Stage 2 (χ²=20.10, P<0.001). Conclusions: Findings suggest two critical periods for increased suicide risk in HD. Although the underlying mechanisms of suicide risk in HD are poorly understood, it is beneficial for health care providers to be aware of periods during which patients may be at an increased risk for suicide. Support from the Huntington's Disease Society of America, Huntington Study Group, and Hereditary Disease Foundation.

P9. Geriatric treatment center: a contemporary model for collaboration between psychiatry and neurologyChristopher M. Filley, David B. Arciniegas, G. Vernon Wood, C. Allen Anderson, Elizabeth Cheney, Charles R. Dygert, Lanier Summerall, Robert B. Goos (University of Colorado Medical School, Denver, CO)

Background: The diagnosis and treatment of individuals with problems involving both psychiatry and neurology have become more sophisticated in recent years, but these advances may be difficult to implement in the modern health care environment. Methods: For the past 16 years, we have employed an inpatient geriatric facility within a state mental hospital to diagnose and treat older persons with complex neuropsychiatric disorders. We present 8 illustrative cases to describe the Geriatric Treatment Center and the procedures used for the care of these challenging patients. Results: All individuals presented with major behavioral dysfunction that could not be managed effectively in other health care facilities. As a result of detailed neuropsychiatric evaluation and behavioral neurology consultation, all had neurologic diagnoses clarified or confirmed as the cause of their presentation. Seven of the patients improved with appropriate treatment, of whom one could be discharged to the community. The eighth patient died and had an autopsy diagnosis of his disease at a nearby academic medical center. Conclusions: This series highlights the value of collaboration between psychiatry and neurology for the evaluation and treatment of older patients with neuropsychiatric problems that are not easily accommodated by many existing health care settings.

P10. Are nonverbal learning disabilities specific types of learning disabilities?Yitzchak Frank, Thomas Pawlcyk, Jessica Seiden (Neurobehavioral Program, North Shore University Hospital, New York University Medical School, New York, NY)

Background: Nonverbal learning disability (NVLD) is a syndrome with features reminiscent of acquired right hemisphere lesions. Subjects have a marked disparity between a higher verbal and a lower performance IQ. Learning difficulties involve mathematical more than reading skills. Visuospatial, attentional, and social communicative skills are impaired. We aimed to determine which aspects of NVLD differentiate it from other types of learning disability. Methods: 70 children and adolescents ages 8 to 15 years evaluated in our program had the following: Wechsler Intelligence Scale for Children–Revised (WISC-R) or WISC-III; a DSM-IV-compatible attention questionnaire (SNAP IV); Achenbach's Child Behavior Checklist and Teacher's Report Form (TRF); a continuous performance test (TOVA), and standardized achievement tests. A designation of NVLD required VIQ scores higher than PIQ scores by at least 12 points and Full Scale IQ score not lower than 85. The following three groups were compared: NVLD, language-based LD, and nonspecific LD. Results: Group differences (P-values reported) were found in the following: WISC-III: Similarities (0.0048), Vocabulary (0.0001), VIQ (0.0001); Achenbach TRF: Total (0.0131), Externalizing (0.032); Teachers' SNAP IV: Inattention (0.019), Hyperactivity/Impulsivity (0.04), Oppositional Defiant Disorder (0.04); TOVA scales: Omission (0.018), Commission (0.009), Response time (0.0069), Variability (0.011); achievement tests: reading comprehension (0.0001), and math (0.0474). These results suggest specific profiles for NVLD.

P11. Specificity of neurological and neurocognitive function in children with attention-deficit/hyperactivity disorderShur-Fen Susan Gau, James R. Merikangas, Kathleen R. Merikangas (Yale University, New Haven, CT)

Background: Previous studies of clinical samples of children with attention-deficit/hyperactivity disorder have suggested that ADHD may be associated with a range of neuropsychologic and neurologic deficits. The objective of this study was to examine the neuropsychiatric correlates of ADHD and their specificity in a nonclinical sample. Methods: Subjects included 203 offspring ages 7 to 18 of parents with substance use disorders; parents with anxiety/depression; and control subjects in a longitudinal family study. Among them, 35 were children with a DSM-III-R diagnosis of ADHD. A comprehensive battery of structured measures of psychiatric, neurologic, and neuropsychologic function was administered blindly by experienced clinicians. Results: After controlling for parental disorders, perinatal factors, comorbid psychiatric disorders, and sociodemographic variables, children with ADHD were more likely than those without ADHD to have lower IQ, shorter spatial memory span, minor physical anomalies (particularly high-arched palate), abnormal MMSE, and abnormal fine motor movements and associated movements on neurologic examination. Integration of these findings revealed that freedom from distractibility (Wechsler) and associated movements were the leading factors discriminating children with ADHD from those with other or no psychiatric disorders. Conclusions: The characteristic neuropsychiatric deficits in these children may inform studies of the neural pathways underlying ADHD and of potential etiologic mechanisms. These findings also suggest the importance of early intensive educational intervention.

P12. EEG findings in women with premenstrual syndromeFrances Hagaman, Mary Jo Fitzgerald, John McWilliams, Margaret Read, Anita Kablinger (Louisiana State University Health Sciences Center, Shreveport, LA)

Background: Changes in the electroencephalogram (EEG) related to the menstrual cycle have been investigated since the 1930s. Authors have found an increase in total power of the alpha range around ovulation (Gautray et al, Acta Eur Fertil 1970; 5–13), and acceleration in alpha rhythms during the luteal phase (Creutzfeld et al, Electroencephalogr Clin Neurophysiol 1976; 128:113–129). Premenstrual syndrome (PMS) represents a cluster of symptoms found in many women during the luteal phase of their cycle. We hypothesized that abnormal EEGs would be found in women who experienced severe PMS. Methods: 81 females between 18 and 40, with regular menstrual cycles, were recruited through newspaper advertisements. All women underwent a psychological interview, completed the Premenstrual Assessment Form (PAF) and the lifetime version of the Schedule for Affective Disorders and Schizophrenia (SADS), and had an EEG. 45 patients completed assessments and met inclusion criteria. Results: Subjects were divided into low, medium, and high PMS groups based on PAF scores. A high number of subjects demonstrated paroxysmal dysrhythmias (PD) despite the severity of PMS symptoms. Subjects who had PD were compared with those having normal EEGs in terms of specific PMS symptoms. Although this study has limitations, the increased prevalence of EEG abnormalities is intriguing.

P13. Amantadine for psychostimulant-resistant attention-deficit/hyperactivity disorder in boysJoseph P. Horrigan, L. Jarrett Barnhill (University of North Carolina, Chapel Hill, NC)

Background: Psychostimulants represent the gold standard of treatment for ADHD. However, they are effective in only 70% to 75% of cases. This study evaluated the safety and efficacy of amantadine hydrochloride for ADHD. Methods: 16 boys (mean age 9.6 years) who attended a residential school participated in this 16-week trial. Each had failed to respond to at least two prior psychostimulant trials. Amantadine was initiated as monotherapy in an open-label manner, at a dose of 100 mg po bid (breakfast, lunch). A further 100 mg was added at 3:00 p.m. after 2 weeks if only a partial response was noted. Results: All 16 boys completed the trial. Few side effects were noted. 13 of the 16 (81%) were classified as positive responders (>30% decrease in the baseline teacher-rated Conners Global Index score and a clinician-generated clinical global impression Improvement score of 1 or 2). The mean effective dose was 262.5 mg/day. Tolerance to amantadine was noted in 7 cases, which was overcome in each instance via a 4-day drug holiday and subsequent rechallenge with the previously effective dose. Conclusions: Amantadine may be a useful addition to the pharmacotherapeutic armamentarium for ADHD, particularly in psychostimulant-resistant cases. Tolerance is a potential problem that appears to be easily managed.

P14. Charles Bonnet syndrome with visual hallucinations of infantile memory: a case reportKazuhisa Maeda, Ryouko Ishikura, Takefusa Nagami, Shu Nukina, Shin-ichi Yoshioka, Ryuzo Kawahara (Tottori University, Yonago, Japan)

Background: Charles Bonnet syndrome (CBS) is a condition in which individuals experience complex visual hallucinations without demonstrable psychopathology or disturbance of normal consciousness. Here we report a case of CBS with visual hallucinations of infantile memory and successful treatment of CBS with using risperidone. Case Report: A 57-year-old woman with a history of depression and diabetes mellitus was admitted to our hospital because of diabetic neuropathy and cataract. Her visual acuity was distinguishing hand movement in the right eye and 0.04 in the left eye. After 7 days of cataract surgery, she experienced visual hallucinations that took the form of a scene of building construction. She maintained insight into the unreal nature of the visions. She realized that she had experienced the scene in childhood. CT and EEG examination revealed no remarkable changes. After 1 week of treatment with 1 mg of risperidone daily, all of her visual hallucinations disappeared. Conclusions: Despite numerous case reports, there are few reports of CBS with visual hallucinations of infantile memory and few successful methods of treatment of CBS. In the case reported here, we successfully treated the condition with the atypical neuroleptic risperidone.

P15. Social cognition deficits in patients with frontal lobe damageLinda Mah, Miriam Courtney Arnold, Jordan Grafman (Cognitive Neuroscience Section, NINDS, Bethesda, MD)

Background: Patients with damage to the frontal lobes frequently exhibit impaired social behavior, despite preservation of language, memory, and perception. This clinical observation has received little systematic evaluation. We studied social cognition in patients with frontal lobe lesions, using the Interpersonal Perception Task (IPT), a standardized method of assessing the accuracy of social judgments regarding videotaped interpersonal situations. Methods: We administered the IPT to patients with lesions involving prefrontal cortex (FLL; n=33) and healthy normal control subjects (n=32). All FLL patients had sustained penetrating head injuries, except for 2 who had aneurysms and 1 with an astrocytoma. Results: FLL subjects performed more poorly on the IPT compared with normal control subjects (P=0.001). They showed specific deficits when making judgments of kinship (P=0.034), determining the veracity of statements (P=0.012), and inferring which individual had competed successfully against the other (P=0.031). FLL patients also tended to overestimate the accuracy of their performance on the IPT compared with the self-ratings of normal control subjects (P=0.06). Conclusions: Patients with damage to prefrontal cortex are impaired in their ability to interpret the nature of social interactions.

P16. Phenomenological and psychopharmacological characteristics of obsessive-compulsive disorder patients with musical obsessionsT. Matsui, H. Matsunaga, K. Ohya, Y. Iwasaki, K. Koshimune, A. Miyata, N. Kiriike (Department of Neuropsychiatry, Osaka City University Medical School, Osaka, Japan)

Background: Intrusive nonsense sounds or music have rarely been reported and discussed in patients with OCD. In practice, drawing a distinction between musical obsessions and musical hallucinations is considered rather difficult. Methods: Clinical features in 2 cases of patients with musical obsessions are presented to discuss phenomenological and psychopharmacological differences from those in patients with musical hallucinations. Results: Despite no covert or systematic compulsive behaviors, the musical symptoms in the present cases are consistent with the phenomenological nature of OCD defined in DSM-IV. In addition, in contrast to previous case reports of musical hallucinations, the present patients failed to respond to neuroleptics but showed significant response to an adequate trial of clomipramine. Conclusions: These symptoms appear to be phenomenologically and biologically distinct from musical hallucinations, especially those characteristic of schizophrenia. However, taking into account the clinical similarities between musical obsessions and musical hallucinations, there may be a number of patients whose musical symptoms could lead to confusion in diagnosis or treatment strategy, and those patients should be treated with caution.

P17. Foreign accent syndrome following a stroke in the corpus callosumDeborah A. Mueller, C. Alan Anderson, Richard L. Hughes, Jill Newcombe (University of Colorado, Denver, CO)

Background: Foreign accent syndrome is a rare constellation of speech anomalies that manifests in a perceived foreign accent. The speech disorder is variable, and only 15 patients have been described in the literature. The etiology of the syndrome is usually stroke or trauma to the left prerolandic motor cortex, the frontal motor association cortex, or the striatum (Kurowski et al, Brain Lang 1996; 54:1–25). Methods: We studied a 54-year-old female who was referred after acute onset of speech difficulties. Results: The patient had a mild receptive aphasia and a French-sounding accent on examination. She had stiff and distorted articulation with frequent hesitancies and paraphrasias. Prosody, praxia, fluency, and volume were all normal. MRI revealed a 1.7-cm hyperintense lesion in the left midportion of the corpus callosum. Conclusions: It was theorized that foreign accent syndrome and other speech disorders could result solely from a disruption of frontal efferent fibers passing across the corpus callosum (Schiff et al, Arch Neurol 1983; 40:720–727). Although the syndrome itself comprises a heterogeneous group of patients with different speech anomalies and localizations, this case may allow us to further elucidate the pathways that are important in producing normal speech patterns.

P18. Meta-analyses of poststroke depression and lesion location: a reappraisalKenji Narushima, Todd Kosier, Robert G. Robinson* (Department of Psychiatry, University of Iowa College of Medicine, Iowa City, IA)

Background: In 1981, we first reported a significant correlation between proximity of left hemisphere stroke lesion to the frontal pole as measured on computed tomography and severity of depression. Since that time there have been numerous replications as well as failed replications, and a meta-analysis in 2000 concluded there was no evidence to support the association of depression with lesion location following stroke. Methods: Journals published from 1981 to 2000 were reviewed. We 1) did a computer-based journal search, 2) used references of reviews and reports, and 3) hand searched the latest key journals. The inclusion criteria required standardized criteria for depression, appropriate imaging, and less than 6 months' interval between stroke and neuropsychoanatomical measurements. Results: 99 reports were identified by the search strategy. 15 reports were eligible for inclusion. Both fixed and random model analyses showed that the association stated above was significant regardless of the side of stroke. Conclusions: This study supported our hypothesis that there is a significant inverse correlation between the severity of depression and the distance of the anterior border of the lesion from the frontal pole up to 6 months following stroke.Support from NIMH Grants MH40355, MH52879, MH53592, and Research Scientist Award MH00163.

P19. Rates of depressive symptoms in presymptomatic Huntington's diseaseC. Nehl, M. Benjamin, J.S. Paulsen,* and the Huntington Study Group (University of Iowa, Iowa City, IA)

Background: It is well established that Huntington's disease is associated with increased levels of depression, although psychopathology in presymptomatic HD has not been adequately studied. Methods: Depressive symptoms, as measured by the depression factor of the Unified Huntington's Disease Rating Scale, were assessed in 2,840 individuals diagnosed with HD and 77 individuals presymptomatic (but having the gene mutation) for HD. All data were obtained from the Huntington Study Group database. Results: A larger proportion of presymptomatic than of diagnosed individuals reported symptoms of sad mood (χ²=83.19, P<0.0001), anxiety (χ²=49.97, P<0.0001), and low self-esteem (χ²=96.93, P<0.0001). Among individuals presymptomatic for HD, 48.6% reported sad mood, 47.3% reported anxiety, and 33.8% reported low self-esteem. In diagnosed individuals, 39.9% reported sad mood, 40.5% reported anxiety, and 24.9% reported low self-esteem. Conclusions: A large proportion of presymptomatic individuals reported depressive symptoms. Further, a significantly larger proportion of presymptomatic individuals showed depressive symptoms compared with individuals having a confirmed diagnosis of HD based on abnormal movements. Further research is required to determine whether differences in rates of depression are attributable to general disease development, adaptation to illness, or reduced awareness of mood.Support from the Huntington's Disease Society of America, Huntington Study Group, and Hereditary Disease Foundation.

P20. Catatonia following acute dopamine depletionR. Nehru, N. Ahuja (Neurobehavior Clinic, GB Pant Hospital, New Delhi, India)

Background: Several reports in the literature demonstrate that catatonia may be induced by neuroleptics, and catatonia may precede or follow the neuroleptic malignant syndrome (NMS), suggesting that catatonia might be a hypodopaminergic state. We present direct evidence in favor of this contention. Case Report: A 51-year-old female Parkinson's disease patient developed multiple catatonic signs preceding a full-blown acute dopamine depletion syndrome (ADDS) with altered sensorium, massive extrapyramidal rigidity, hyperthermia, autonomic disturbance, and markedly elevated serum creatine phosphokinase, following planned reduction of levodopa therapy for dopa-dyskinesias. She was still recovering with restoration of levodopa dosage and addition of bromocriptine when she again developed catatonia following inadvertent reduction in her dosage. She never received any dopamine antagonists and did not receive lorazepam or ECT for treatment of catatonia. Catatonia appeared twice following withdrawal of levodopa in this case. It resolved on both occasions, along with the ADDS, following restoration of levodopa and addition of bromocriptine. Conclusions: Observations suggest that both NMS (functionally equivalent to ADDS) and catatonia are hypodopaminergic states, and that both lie on the same spectrum of disorders. Hypodopaminergia may be the fundamental neurochemical denominator in catatonia, along with secondary or consequential perturbations of other transmitter systems.

P21. Hypersexuality following right pallidotomy for Parkinson's diseaseSusan M. O'Connor, Gerald T. Lim, Mario F. Mendez* (University of California at Los Angeles, Los Angeles, CA)

Background: Sexual behavior is mediated by the hypothalamic preoptic nuclei and the septal nuclei and is modulated by frontal and temporal lobes. The role of the globus pallidus (GP) in sexual expression is less clear. We report a patient with marked hypersexuality following right pallidotomy for intractable Parkinson's disease. Case Report: Immediately after right pallidotomy, a 59-year-old man began demanding of his wife of 41 years coital activity up to several times per hour. He masturbated constantly and nearly bankrupted his family from spending on prostitutes, strippers, and pornography. He propositioned his wife's female friends and was allegedly sexually inappropriate with his granddaughter. Review of his PD medications did not disclose a correlation with his hypersexuality. On examination, the patient endorsed increased sexual interest but was not distressed by it. Mental status was otherwise normal. Pallidotomy relieved his tremor and gait disturbance, but he remained with mild rigidity and bradykinesia. MRI showed a discrete lesion in the right GP. His hypersexuality responded to valproate therapy. Conclusions: This patient developed hypersexuality after right pallidotomy. Rauch et al. (Psychiatry Res 1999; 91:1–10) report increased blood flow in ventral GP during sexual arousal, further suggesting that the GP participates in the neurocircuitry of sexual behavior.

P22. The Charcot-Willibrand syndrome from anoxic encephalopathyAlireza Pakkar, Gerald T. Lim, Mario F. Mendez* (University of California at Los Angeles, Los Angeles, CA)

Background: The loss of visual imagery is known as the Charcot-Willibrand syndrome. Patients with this condition have lost the ability to represent images visually in their mind. We report a patient with this rare syndrome. Case Report: A 69-year-old man with depression attempted suicide by hanging himself. He was found unconscious and transferred to intensive care with presumed anoxic brain injury. After recovering, he complained of inability to visualize objects, faces, or routes and could not remember details of complex figures and faces, including his own face. He described his mind as visually blank and complained of inability to see himself in his dreams. On examination, mental status was normal. Memory, language, and executive functions were also normal. On visual testing, he could not describe from memory the visual features of complex images, objects, or faces but had no problem recognizing and matching these images. There was no evidence of visuoperceptual deficits, visual agnosia, visuospatial impairment, prosopagnosia, or environmental disorientation. MRI of brain showed mild cerebral and cerebellar atrophy. Conclusions: This rare syndrome of defective revisualization results from presumed parietal lobe injury. This patient illustrates a postanoxic etiology for the Charcot-Willibrand syndrome and its dissociation from other visuoperceptual deficits.

P23. Neurobehavioral disorders in pediatric epilepsyGustavo J. Rey, Leanne Shapiro, Trevor J. Resnick, Michael Duchowny, Prasanna Jayakar (Neuroscience Center, Miami Children's Hospital, Miami, FL)

Background: Existing data indicate a significantly higher frequency of behavioral disorders in children with epilepsy relative to the general population. Yet there is a dearth of pediatric studies assessing the relationship between clinical-neurological? variables and psychiatric comorbidity in this group. Methods: We examined 22 children referred to a tertiary epilepsy center. The sample had a mean age of 12.8 years (SD=3.9) and was equally divided as to gender. Average age at onset was 4.0 years (SD=3.2), and mean seizure frequency was 14.22 episodes/week. Some of the subjects had intractable epilepsy and were being examined for epilepsy surgery candidacy. Results: 45% of the sample had a Full Scale IQ within the mentally retarded range. There was a high prevalence of behavioral disorders, with 40% demonstrating clinically significant social/interpersonal problems and 32% presenting symptoms of ADHD. Age at onset, seizure frequency, presence/absence of an identifiable lesion, and lateralization of EEG findings were not related to intellectual and behavioral status. Conclusions: Our study revealed a high prevalence of intellectual deficiencies and behavioral disturbances in a heterogeneous pediatric epilepsy sample. Larger studies using more homogeneous samples are needed to understand the neurological substrates of these associated psychiatric disorders in epilepsy.

P24. Depressive symptoms in Parkinson's disease are underrecognizedIrene H. Richard, Anne W. Justus, Eric Caine, Roger Kurlan (University of Rochester, Rochester, NY)

Background: Depression in Parkinson's disease is associated with increased disability (Liu et al, J Neurol 1997; 244:493–498) and reduced quality of life (Kuopio et al, Mov Disord 2000; 15:216–223). Methods: 100 PD patients and 42 spouses were invited to answer questions about their psychiatric history and medications and to complete the 13-item Shortened Beck Depression Inventory (S-BDI) and the Geriatric Depression Rating Scale (GDS). Results: 76 patients and 15 spouses participated. 40 (53%) of the patients, but no spouses, scored within the mildly depressed range or higher on the S-BDI (>7) and/or the GDS (>9). 24/40 (60%) denied ever being diagnosed with depression or treated with antidepressants. Seven patients (9%) scored within the severely depressed range on the S-BDI (>10) and moderate–severe range on the GDS (>20). Only 1/7 reported a history of depression diagnosis and was taking an antidepressant. Conclusions: It will be important to confirm diagnoses of depression with face-to-face interviews, but these results suggest that many PD patients suffer from unrecognized and potentially treatable depressive symptoms.Support from NINDS and NARSAD.

P25. Are delirium and dementia different?Paula T. Trzepacz, Dinesh Mittal, Rafael Torres, Kim Kanary, John Norton, Nita Jimerson (University of Mississippi Medical Center, Jackson, MS; Lilly Research Laboratories, Indianapolis, IN)

Background: This study compared phenomenological differences between DSM-IV delirium (DEL) and dementia (DEM), using two validated scales specifically designed for use in delirium. Methods: Ratings on the Delirium Rating Scale-Revised-98 (DRS-R-98) were done by psychiatrists blind to diagnosis and to Cognitive Test for Delirium (CTD) scores. Items on the DRS-R-98 and CTD were compared by using nonparametric tests between DEL (n=24) and DEM (n=13) groups. Results: Overall scores on the CTD (P<0.05) and DRS-R-98 (P<0.001) were worse in DEL than DEM. CTD items for orientation, memory, and vigilance did not differ. Comprehension was the only CTD item significantly worse in DEL than DEM (P=0.009), although attention span showed a trend (P=0.076). CTD comprehension correlated with DRS-R-98 total score (r=–0.58, P=0.003) in the DEL group. On the DRS-R-98, DEL had significantly higher scores (more impairment) for sleep-wake cycle disturbance (P<0.001), perceptual disturbance (P=0.006), thought processes (P=0.003), lability of affect (P=0.009), motor agitation (P<0.001), attention (P=0.001), and visuospatial ability (P<0.001), as well as for the three diagnostic items: temporal onset of symptoms (P<0.001), fluctuation of symptoms severity (P<0.001), and physical disorder (P<0.001). There were no significant differences for DRS-R-98 delusions, language, motor retardation, orientation, short-term memory, or long-term memory, however. Conclusions: Both the Cognitive Test for Delirium and DRS-R-98 found comparable deficits in orientation and memory in DEL and DEM but detected more impaired attention in DEL. These findings support DSM criteria for cardinal attentional impairment in delirium, but not for memory deficits unique to dementia. All three DRS-R-98 diagnostic items distinguished DEL from DEM, supporting empirical literature. DRS-R-98 identified more distinguishing symptoms than previous reports (Liptzin & Levkoff), which could highlight underlying neural differences between delirium and dementia. Delusions, language, motor retardation, orientation, and short- and long-term memory were comparably impaired in DEL and DEM.Support from the Mental Illness Research Education and Clinical Center, Veterans Integrated Service Network 16 (MIRECC-VISN 6), Department of Veterans Affairs.

P26. An investigation of the psychological functioning and suicide risk among relapsing-remitting multiple sclerosis patientsYakov Vardi, Zoli Zlotogorski, Yoram Finkelstein (Shaare Zedek Medical Center, Jerusalem, Israel)

Background: Research has reported depressive symptomatology and affective lability among multiple sclerosis patients (Fisk et al, Can J Neurol Sci 1998; 23:230–235). Of particular concern are the reported higher suicide attempts for this population. Methods: 28 nonmedicated patients with relapsing-remitting MS (Expanded Disability Status Scale mean ∼3.5) in the remitting phase were administered projective tests. The 16 women and 12 men who participated in the study had a mean age of 39 and 14 mean years of education. Exner's comprehensive system of administration and scoring was employed to ensure reliability and validity. Results: The MS sample differed significantly from the normative sample on nearly all personality variables. Our findings suggest that MS patients may suffer from a thought disorder, not previously reported, that is comorbid with the affective disorder. More important, the MS population evidenced a higher suicide risk on the suicide potential measure, which seems to be related also to the thought disorder. SPECT study of these patients showed hypoperfusion of the left prefrontal region. Conclusions: Higher risk of suicide in MS patients is related to thought disorder no less than to the depressive state.

P27. Episodic supernumerary phantom limb due to epileptic disruption of the neural representation of the body imageAaron Weisbord, Jeffrey L. Saver,* Megan C. Leary (University of California at Los Angeles, Los Angeles, CA)

Background: The phantom limb syndrome is a common sequela of limb amputation. Rarely reported is the sensation of having extra appendages, the supernumerary phantom limb syndrome or pseudopolymelia. Case Report: A 78-year-old female complained of 10 days of episodes of tingling and numbness in the left arm, followed by the subjective sensation of a second left arm growing out of middle of her chest wall. When she moved her true left arm, the phantom left arm moved identically. Episodes lasted less than 20 s and occurred 4 to 5 times a day. Shortly after hospitalization, the patient suffered a left radial fracture requiring cast placement. During two further spells, she reported the sensation of a cast on the supernumerary limb. Diffusion-weighted MR imaging showed a small right parietal cortical infarct. EEG revealed sharp spike activity, involving the left temporal region, and right temporal slowing. Spells resolved with initiation of phenytoin. Conclusions: In reported cases, most supernumerary phantom limb patients present with sustained pseudopolymelia in the setting of hemiplegia, sensory impairment, and anosognosia, in association with a large right hemispheric stroke. Ours is only the second case reported of recurrent, episodic supernumerary phantom limb. Our case suggests that transient supernumerary phantom limb sensations can be a manifestation of focal seizures. The case implicates in the genesis of pseudopolymelia the parietal systems responsible for sensorimotor integration and internal representation of body image.

P28. Sexual fantasy in Parkinson's diseaseMi Yu, David Roane, John Rogers, Megan Fleming, Peter Homel (Beth Israel Medical Center, New York, NY)

Background: Parkinson's disease is associated with sexual dysfunction in up to 80% of patients. This study examined the biopsychosocial correlates of sexual functioning in PD, considering the effect of PD on the different phases of sexual response. Methods: 13 male and 4 female PD patients, ages 53–85, were evaluated with their partners. Measures included the Derogatis Interview for Sexual Functioning (an assessment of function across the sexual response cycle), Unified Parkinson's Disease Rating Scale, Ham-D and Ham-A, Dyadic Adjustment Scale, Social Support Questionnaire, and assessment of autonomic dysfunction. Results: Subjects demonstrated substantial impairment in the arousal, orgasm, and sexual drive phases, with most patients scoring below the first percentile on these Derogatis Interview subscales. Sexual fantasy scores, with a median percentile of 29%, allowed for meaningful correlation analysis. Fantasy scores failed to correlate with age, previous sexual experience, or current sexual behavior. Sexual fantasy tended to correlate negatively with anxiety (r=0.424, P=0.09) and positively with duration of PD (r=0.715, P<0.01). Conclusions: Sexual fantasy is relatively preserved, compared with other aspects of sexuality, in PD. The use of fantasy may actually increase with disease progression, suggesting that clinicians should attend to sexual concerns throughout the course of PD.

Dementia

P29. Does cognitive recovery after treatment of poststroke depression last? A two-year follow-up of cognitive function associated with poststroke depressionKeen-Loong Chan, Robert G. Robinson, James T. Kosier (University of Iowa, Iowa City, IA; Tan Tock Seng Hospital, Singapore)

Background: Cognitive impairment is common after stroke and may be caused by poststroke depression and/or structural damage to the cerebrum. We examined the changes in cognition over 2 years after remission of poststroke depression (PSD). Methods: 12 patients with PSD and cognitive impairment who were selected on the basis of early and sustained remission of their depression were compared with 12 nondepressed stroke patients. Mood and cognition were followed up over 2 years, using the Ham-D and MMSE. Results: In patients with early and sustained remission of depression, there was rapid and early improvement of cognitive function, which was maintained over 2 years. For depressed patients, the initial MMSE score of 22.5±3.8 improved to 26.7±3.0 at 3 months and was 25.6±3.3 at 2 years. The magnitude of cognitive impairment due to depression was estimated as –3.4 points on the MMSE. Conclusions: Cognitive function, once improved after remission of PSD, is likely to remain stable over the next 2 years in the absence of subsequent reinjury to the central nervous system. Cognitive impairment due to poststroke depression is reversible and can be quantified separately from irreversible cognitive impairment due to other aspects of stroke. Support from NIMH Grants MH40355, MH52879, and MH53592.

P30. Open-label trial of guanfacine for symptomatic treatment of frontotemporal dementiaTiffany W. Chow (Rancho Los Amigos National Rehabilitation Center/University of Southern California Alzheimer's Disease Center, Los Angeles, CA)

Background: Children with attention deficit disorder (ADD) have a frontal dysexecutive syndrome similar to that of most patients with frontotemporal dementia (FTD). Treatment with alpha₂ agonists, clonidine or guanfacine, improves attention and concentration ability. Guanfacine has fewer side effects in children with ADD than clonidine. It is possible that guanfacine could also address frontal dysexecutive syndrome in older adults with FTD. Methods: 14 adults with FTD syndromes (frontal lobe degeneration and primary progressive aphasias) consented to an open-label trial of guanfacine. Over a 3-month period, subjects took a maximum dose of 1 mg daily. A continuous performance task (A-test), digit span forward and reverse, Stroop C and C-W, Trails A and B, and the Neuropsychiatric Inventory were administered at baseline, 1-month, and 3-month time points. Results: Subjects with moderate to advanced illness (<111/112 on the Stroop C simple reading task) did not benefit from guanfacine at the 1-mg dose. Improvement in frontal executive function was equivocal, and caregiver distress worsened among those subjects in early to moderate stages. Conclusions: Evidence-based treatment for patients with FTD is in great demand from patients and their families. Multicenter trials using different agents or using a higher dose of guanfacine in early stages of FTD may shed more light on treatment options.

P31. A review of the memory stimulation programs in Alzheimer's diseaseÉric Grandmaison, Martine Simard (Université de Moncton, Moncton, Nouveau-Brunswick, Canada)

Background: The goal of this work was to qualitatively review the efficacy of the memory stimulation programs used in Alzheimer's disease. Methods: MEDLINE and PsycINFO searches were done using the following keywords: memory remediation; memory remediation and dementia; cognitive remediation; cognitive remediation and dementia (and then with “stimulation” instead of “remediation”); vanishing cues; spaced retrieval; errorless learning; cue utilization and AD; visual imagery and AD. Results: Six controlled and five randomized controlled studies as well as four studies with multiple single-case experimental design were reviewed. N per group ranged from 2 to 33 subjects. Visual imagery, the errorless learning and dyadic approaches, spaced-retrieval techniques, encoding specificity with cognitive support at retrieval, and external memory aids were the memory stimulation programs used alone or in combination in AD. Studies using the space-retrieval techniques (alone or in combination with the errorless learning approach) and the dyadic approach showed the best positive effects and long-term maintenance of the gains up to 6 and 8 months post baseline. Conclusions: Preliminary evidence suggests that the errorless learning, dyadic approach, and spaced retrieval techniques, used alone or in combination, are efficacious to stimulate memory in AD. More crossover, randomized, placebo-controlled studies, using larger sample of subjects, are needed.Support from the Jeanne et J.-Louis Levesque Doctoral Award (E.G.).

P32. Early results of the REVEAL study: risk evaluation and education for Alzheimer's diseaseR.C. Green, N. Relkin, P. Whitehouse, T. Brown, S. LaRusse, M. Barber, L. Farrer, L. Cupples, S. Post, L. Ravdin, D. Cisewski, H. Katzen, S. Sami, A. Sadovnick, J. Davis, K. Quaid, J. Woodard (Boston University School of Medicine, Boston, MA)

Background: The REVEAL study is a randomized controlled trial examining the impact of providing risk assessment, including apolipoprotein E (APOE) genotyping, to asymptomatic adult children of persons with Alzheimer's disease. Methods and Results: 242 REVEAL participants have been enrolled, 181 have attended the introductory educational session, 133 have proceeded through neuropsychology screening and blood draw, 113 have received results disclosure, and 102 have been seen for one or more follow-up visits. Enrolled individuals who chose not to continue through the results-disclosure step cited insurance concerns, and some changed their mind about wanting to learn risk or genotype information. Of those seen for results disclosure, 81 were randomized to receive a personalized risk estimate based on both family history and APOE genotype information. The remaining 32 participants (control group) were randomized to receive a risk estimate based on family history information only. Of those participants who have received APOE results thus far, 37 had a 3/3 genotype, 35 had 3/4, 4 had 2/3, 4 had 4/4, and 1 had 2/4. Participants' reactions to personalized risk estimates were varied and included being upset, having preconceived risk beliefs confirmed, and/or feeling reassured by the information presented. Outcome variables include measures of anxiety, depression, and satisfaction with the risk assessment and counseling protocol, as well as real-world decisions to change retirement planning or insurance coverage. Conclusions: Disclosures of APOE genotype and AD risk assessment have been well tolerated thus far. Additional results of early analyses will be presented.Support from the National Human Genome Institute and NIA.

P33. Differential performance on the Benton Judgment of Line Orientation Test in dementia with Lewy bodies and Alzheimer's diseaseMartine Simard, Robert van Reekum, David Myran, Michel Panisset, Sandra E. Black, Morris Freedman (Université de Moncton, Moncton, Nouveau-Brunswick, Canada)

Background: Reports of differential impairments on visual-construction tasks in dementia with Lewy bodies (DLB) and Alzheimer's disease (AD) are sometimes controversial, and visual-perceptual data are lacking (Simard et al, J Neuropsychiatry Clin Neurosci 2000; 12:425–450). The goal of this study was to assess the visual-perceptual performance of subjects with DLB and AD. Methods: The Benton Judgment of Line Orientation Test (BJLO) was administered to 8 DLB, 13 AD, and 8 control subjects. An analysis of error types (Ska et al, J Clin Exp Neuropsychol 1990; 12:695–702) was afterwards applied to the results of the BJLO with QO1, QO2, QO3, QO4 (visual attention) errors, as well as VH, IQOV, and IQOH (visual-spatial perception) errors. Results: A MANOVA showed significant differences between the three groups on the BJLO total score (F=4.09, df=2,26, P<0.05) and the number of errors for VH (F=5.45, df=2,26, P<0.05) and IQOH (F=3.83, df=2,26, P<0.05). The t-test analyses demonstrated that subjects with DLB made significantly more VH (P<0.05) and IQOH (P<0.05) errors than subjects with AD. Compared with control subjects, both patient groups had inferior performance on the BJLO total score. Conclusions: Subjects with DLB have more severe visual-perception (VH and IQOH errors) impairments than subjects with AD.Support from Alzheimer Society of Canada.

POSTER SESSION II—TUESDAY, MARCH 12

Neuropsychology

P34. Coping strategies moderate the relationship between cognitive dysfunction and depression in multiple sclerosis patientsPeter A. Arnett, John J. Randolph, Pamela J. Freske (Pennsylvania State University, University Park, PA)

Background: Cognitive dysfunction has been shown to be inconsistently related to depression in multiple sclerosis patients. This inconsistency may be due in part to the influence of moderating factors heretofore unexamined. Patients with speeded attentional deficits commonly associated with MS may be at greatest risk for depression when they also use maladaptive coping strategies, such as those involving disengagement. The current study was designed to test this hypothesis. Methods: Correlations between three speeded attentional measures shown to be most sensitive to depression from our prior work (Arnett et al, Neuropsychology 1999; 13:434–436) and a measure of evaluative and mood symptoms of depression (Chicago Multiscale Depression Inventory) were examined separately in MS groups high (upper 30% of sample, n=14) and low (lower 30% of sample, n=15) on a “Disengagement” coping factor defined from the COPE. Results: In the high-disengagement coping group, correlations were significant between the depression index and all three speeded attentional measures (P-levels from 0.05 to 0.005). None of the same correlations were significant in the low-disengagement coping group. Conclusions: Our data show that disengagement coping moderates the relationship between depression and speeded attentional dysfunction in MS. Speeded attentional dysfunction may result in depression in MS only when patients also use maladaptive coping strategies involving disengagement.

P35. Relationship between depression and attentional/memory processes in multiple sclerosisSamuel T. Gontkovsky, Robin Aupperle, Fatima Shelton, William W. Beatty (University of Oklahoma Health Sciences Center, Oklahoma City, OK)

Background: Adams et al. (Arch Clin Neuropsychol 2001; 16:605–618) recently postulated an indirect association between psychopathology and memory functioning in heterogeneous psychiatric and brain-damaged patient samples, with attentional processes serving as a mediating variable. The present investigation examined this relationship in patients with chronic multiple sclerosis. Methods: Participants were 44 patients (mean±SD: age, 50.6±8.7 years; education, 14.5±2.1 years) diagnosed by neurological examination with MS (mean disease duration, 14.0±8.6 years) who underwent voluntary neuropsychological examination. The Chicago Multiscale Depression Inventory (CMDI) was used to assess depressive symptomatology. The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) was employed to measure cognitive functioning. Results: Multiple regression analyses revealed neither the evaluative scale nor the vegetative scale of the CMDI was predictive of attentional or memory functioning as measured by the RBANS. Scores on the mood scale of the CMDI, by contrast, were found to be predictive of RBANS immediate and delayed memory but not attentional processes. Conclusions: Memory processes are directly associated, although only modestly, with depressive mood in patients with chronic MS. A relationship between attentional skills and depressive symptomatology is not evident in this population.

P36. Attentional impairment in recurrent major depressionÅ. Hammar, A. Lund, K. Hugdahl (University of Bergen, Norway)

Background: The aim of this study was to investigate cognitive functions, with a focus on automatic and effortful processing, in depressed patients at two test occasions: at inclusion and after 6 months. All patients were investigated within a neurocognitive experimental setting based on Treisman's visuals search paradigm (Treisman & Gelade, Cognit Psychol 1980; 12:97–136). The hypothesis is that depressed patients will show a normal performance (reaction time; RT) when automatic processing is required, but show an impaired profile when effortful processing is demanded. Methods: 20 patients with recurrent major depression according to DSM-III-R criteria and a Ham-D score of >18 were included in the study. Ham-D mean score was 21.6 at inclusion and 11.4 after 6 months. The control group was matched by age and gender. Results: The depressed patients performed equally to the control group on trials requiring automatic processing. The depressed patients showed no improvement in reaction time from test occasion 1 to test occasion 2 on any condition. The control group showed a significant improvement in RT on the second test compared with the first. Further, the control group showed a significantly faster reaction time compared with the depressed patient group on all trials requiring effortful processing at both test occasions. Conclusions: The results indicate that the depressed patients have impaired performance on effortful, but not automatic, information processing, and the impairment seems to persist for at least 6 months.

P37. Cognitive impairments associated with hypothermic-induced platelet aggregation (HIPA) following coronary artery bypass graft surgeryRamona O. Hopkins, Matthew M. Hall, Kenneth A. Solen (Brigham Young University, Provo, UT)

Background: Hypothermic-induced platelet aggregation (HIPA) occurs in platelet-rich plasma and differs in mechanism from chemical- or shear-induced platelet aggregation. We have observed HIPA in whole blood at 32° C with a maximum response at 24° C and have identified that 32% of individuals exhibit some HIPA and 14% exhibit extensive HIPA. Methods: We enrolled 39 subjects (32 males, 7 females) with a mean age of 65±7 years and education of 14±3 years, who underwent hypothermic (32°–28° C) coronary artery bypass graft surgery (CABG). All subjects had cognitive testing before and 4–6 days after surgery. The presurgical IQ scores were VIQ=106±8.7, PIQ=105±12.3, and FSIQ=106±11.2. A blood sample from each subject was screened in vitro to identify HIPA propensity. Neuropsychological testing included the Rey-Osterrieth Complex Figure (Roc), Pre-Op Digit Symbol/Post-Op Symbol Digit (DS/SD), Rey Auditory Verbal Learning (RAVL), Trail Making Test Parts A and B (TM), and verbal fluency (VF). Results: Preliminary results indicated impaired postoperative cognitive performance for all patients on DS/SD, TM, and RAVL, compared with preoperative cognitive test scores (P<0.01). Patients with HIPA showed a greater decrease in cognitive function (P<0.01) in TM and VF compared with non-HIPA patients. Conclusions: We found that hypothermic CAGB surgery reduces cognitive function and that patients with HIPA have greater cognitive impairments compared with non-HIPA patients.Support from the LDS Hospital Deseret Foundation.

P38. Thalamic acalculiaGerald T. Lim, Nora C. Papasian, Margaret Swanberg, Mario F. Mendez* (University of California at Los Angeles, Los Angeles, CA)

Background: “Acalculia” refers to impairments in mathematical abilities from brain injury. Classifications of acalculia usually include alexia and agraphia for numbers, spatial acalculia, and anarithmetia. More recent formulations suggest more extensive subtypes, including a number-fact retrieval system (left angular gyrus) and a calculation system (bilateral parietal regions). In addition, frontal regions are involved in working memory and organization for calculating. We report a patient with acalculia from a left thalamic stroke and analyze his deficits. Case Report: A 61-year-old carpet installer developed acute acalculia with difficulty multiplying and using complex mathematical concepts. His mental status was otherwise intact. Originally, the patient performed poorly on multiplication with sparing of addition and subtraction. In subsequent evaluations he improved on the EC301 Battery (Deloche et al, J Int Neuropsychol Soc 1996; 2:177–180), but he was unable to analyze multistep numerical problems and organize the steps needed to solve them. He remained unable to do the calculations required to install carpets. MRI scan revealed a new left thalamic lacune. PET was normal. Conclusions: Thalamic acalculia, previously reported from a left tuberothalamic infarction (Bogousslavsky et al, Stroke 1986; 17:434–441), may result from CQ difficulty integrating the different steps in a complex mathematical problem.

P39. Exaggeration of distress and dissimulated memory failureRichard I. Naugle, Tara T. Lineweaver, Emily Walton (Cleveland Clinic Foundation, Cleveland, OH)

Background: Depression and malingering are both potential confounds in memory assessment. However, the relationship between these factors is poorly understood. Methods: The Victoria Symptom Validity Test (VSVT) is a two-option forced-choice measure designed to detect feigned memory disorders; scores that are significantly below chance levels indicate intentionally substandard performances. The VSVT and the Personality Assessment Inventory (PAI) were administered to 175 patients in the course of their neuropsychological evaluations. Participants were classified as Valid, Questionable, and Invalid responders based on their VSVT performances. The three groups' mean scores from the PAI were compared in ANOVAs. Results: Respondents obtaining VSVT scores in the Questionable and Invalid ranges tended to have higher Depression scores than those in the Valid group (F=6.53, P<0.01). However, comparison of mean validity scale results revealed that those individuals with questionable or invalid VSVT performances also produced higher mean scores on the PAI Negative Impression Management scale (F=4.03, P<0.05). Conclusions: Results indicate that patients in this sample who were feigning or exaggerating memory compromise were also more likely to exaggerate their psychological symptomatology, calling into question the clinical validity of self-reports of depression in patients who are suspected of dissimulating memory deficits.

P40. Mediation of verbal memory deficit by organizational strategy in Parkinson's diseaseRobert M. Roth, Laura A. Flashman, Andrew J. Saykin, Heather A. Wishart, Henry J. Riordan, David W. Roberts (Dartmouth Medical School, Lebanon, NH)

Background: Previous research has shown verbal learning and memory deficits in Parkinson's disease. Several authors have suggested that these deficits are due in part to impaired internal generation of organizational strategies. The present study employed statistical mediation modeling to determine whether verbal learning and recall deficits in PD can be accounted for by deficient use of semantic organization during encoding. Methods: Participants included 23 nondemented patients with Parkinson's disease and 22 healthy control subjects matched for age, education, sex, and handedness. Participants completed the California Verbal Learning Test (CVLT) and a semantic category fluency task. Results: Test results revealed lower performance by the PD group for words recalled during learning and delayed recall trials; lesser use of semantic clustering during encoding; and lower semantic fluency. Regression analyses showed that semantic clustering during encoding mediated group differences in learning and delayed recall. Hierarchical regression analysis indicated that the relationship between semantic clustering and memory measures in the PD group was not accounted for by impaired access to or size of semantic stores. Conclusions: The pattern of findings provides support for the hypothesis that semantic organization accounts in part for verbal learning and memory deficits in PD. The findings are consistent with other studies observing this relationship in disorders with frontostriatal system involvement.

P41. Persistent focal retrograde amnesia: functional or organic?Ronald E. Saul (University of California at Los Angeles, Los Angeles, CA)

Background: Whether persistent focal, i.e., isolated, retrograde amnesia (RA) is a valid neurological entity or invariably psychogenic in origin remains controversial. (Kopelman, Cognitive Neuropsychology 2000; 17:585–621) Case Report: A 30-year-old mechanic complained of headache and other postconcussive symptoms following mild head trauma. Post angiography, he suddenly developed a dense RA, extending back to early childhood, but ongoing memory was unaffected. A brain CT scan and several EEGs were normal. Except for the RA, there were no persistent focal neurological or cognitive deficits on subsequent medical and neuropsychological evaluations. When personally examined 2 years later, he claimed an isolated and extensive RA, encompassing memory of specific events, general knowledge, and learned skills. Attention, recent memory, new learning, and language were preserved. He initially did not recognize his family, his own face in the mirror, objects, or common foods and “had to relearn everything,” e.g., the meanings of words, how to tie his shoes and drive a car. Conclusions: Normal brain imaging and the absence of anterograde amnesia, coupled with a significant loss of semantic and procedural memory, supports psychogenic causation. The differential diagnosis of focal RA and the possible neural basis of such retrieval deficits will be discussed.

P42. Cognitive functioning in patients with type 2 diabetes examined by using the Cognistat (NCSE)Robert L. Yuspeh, Daniel L. Drane (Johns Hopkins University School of Medicine, Baltimore, MD)

Background: Previous research has demonstrated that persons with type 2 diabetes display deficits in the areas of learning, memory, and attention. Methods: The present study examined cognitive functioning in a group of patients with type 2 diabetes (n=16) with a demographically matched group of healthy control participants (n=21), using the Neurobehavioral Cognitive Status Exam or Cognistat (also known as the NCSE). ANCOVA applied in multivariate models was used to compare the type 2 diabetes group with the control group on a newly devised total Cognistat score as well as scores on individual subscales controlling for history of cardiovascular disease, cerebrovascular disease, and depression. Results: Statistically significant differences were noted on total Cognistat score (P=0.03; diabetes accounted for 42% of the variance on this measure), as well as on the Orientation subscale (P=0.04; diabetes, 35% of variance) and the Memory subscale (P=0.006; diabetes, 40% of variance); type 2 diabetes patients performed significantly lower than control subjects. Conclusions: In findings consistent with previous research, type 2 diabetes patients displayed deficits in learning and memory on a commonly used cognitive screening measure. Implications of these results will be presented, as will recommendations for future research.

Schizophrenia

P43. Compliance in patients with schizophrenia and comorbid medical illnessCesar E. Munoz-Carbone, Warren T. Jackson,* Robert M. Savage, Jacqueline M. Feldman, William G. Ryan, L. Jan Findlay, Andrea J. Viegas, Roberta Sokol May (University of Alabama at Birmingham, Birmingham, AL)

Background: Many patients with schizophrenia also suffer from chronic medical conditions, such as hypertension, diabetes, and chronic respiratory problems, that are often unrecognized and underdiagnosed (Jeste et al, Schizophr Bull 1996; 22:413–430). The present study evaluated differential compliance with psychiatric and medical treatments and examined the relationship between compliance, patients' neurocognitive status, and their level of understanding about schizophrenia and comorbid medical disorders. Methods: A consecutive sample of 43 inpatients with schizophrenia or schizoaffective disorder consented to comprehensive psychological and neuropsychological evaluation while hospitalized and to 1-year retrospective and 6-month prospective chart reviews. Medical and psychiatric treatment compliance was independently rated by faculty psychiatrists on the basis of information distilled from detailed chart reviews. Results: No significant differences were found between medical and psychiatric treatment compliance ratings in either the retrospective or prospective time intervals. Prospective psychiatric compliance was significantly better than retrospective; however, there was no difference between medical compliance ratings in the retrospective and prospective intervals. Neither patients' neurocognitive status nor their understanding of illnesses correlated significantly with prospective ratings of compliance. Conclusions: Compliance with psychiatric treatment during the 6 months after inpatient hospitalization was best predicted by compliance during the year prior to hospitalization, but it accounted for only about 18% of the variance. Patients were not differentially compliant with their psychiatric and medical treatment programs.Support from Hoechst Marion Roussel 1998 Quality Care Research Fund.

P44. Neurological soft signs as predictors of specific cognitive functions in first-episode schizophreniaDaniel Schuepbach, Matcheri S. Keshavan, Richard D. Sanders (Western Psychiatric Institute and Clinic, Pittsburgh, PA)

Background: Neurological soft signs (NSS) have become an important research instrument for the assessment of subtle neurological impairments in schizophrenia. However, there are very few studies in first-episode schizophrenia investigating the association of NSS with other relevant markers of the disease, such as cognitive deficits. Methods: 59 neuroleptic-naive first-episode schizophrenia patients and 51 healthy individuals, matched for age, gender, and parental socioeconomic status, were included. All subjects were assessed by use of clinical ratings (SANS and SAPS negative and positive symptom scales), a comprehensive neuropsychological battery, and a modified Neurological Evaluation Scale (NES). Results were analyzed with principal component analysis for NES and stepwise multiple linear regressions for neuropsychological test performance, using single NES factors as predictors and controlling for sociodemographic and clinical covariates. Results: Patients with schizophrenia showed decreased performance on NSS (P<0.001) and neurocognition (P<0.001). The cognitive-perceptual domain of NES predicted specific neuropsychological functions measuring attention (WAIS-R Digit Span: ΔR=0.14, P<0.01), executive function (perseverative errors of WCST: ΔR=0.09, P<0.05), and language skills (FAS verbal fluency: ΔR=0.70, P<0.05). However, there was no association between motor dimension of NES and neuropsychological tasks. In healthy control subjects, NSS predicted verbal and nonverbal memory functions. Conclusions: These results suggest that parts of the NES can be used to reliably predict mainly prefrontal dysfunction in schizophrenia. The findings also imply that there are specifically impaired cognitive functions in schizophrenia.Support from a Western Psychiatric Institute and Clinic research fellowship; Novartis Foundation and Swiss Foundation for Medical Biological Grants; and NIMH Grants MH45203 and MH45156.

Basic Neuroscience

P45. Critical Endeavors: a neuroanatomy board gameRobin A. Hurley, Katherine H. Taber (Mental Health Care Line, Houston Veterans Affairs Medical Center, Houston, TX; Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine Department of Radiology and Herbert J. Frensley Center for Imaging Research, Baylor College of Medicine, Houston, TX)

Students retain and integrate new information much better when they are actively engaged in the learning process. Promoting active engagement is particularly difficult in medical education because the information density is high and time is limited. One approach is to provide students with materials to be studied independently, then use in-class time for exercises in which students actively utilize the information. We have created a board game (Critical Endeavors) for the purpose of helping psychiatry residents learn functional neuroanatomy. This topic is difficult for them because many structures, anatomic relationships, and syndromes must be mastered. The residents move around the board by the roll of dice, landing on squares that require them to answer questions (provided on cards) at various levels of difficulty, gaining or losing money based upon their answer. Once an answer is given they review the correct answer (also provided on each card). Residents become quite actively engaged in the process, promoted by the use of humor in many of the questions, and learn their neuroanatomy at least as well as when taught in a lecture format. The gameboard, sample questions, and instructions for play will be presented as well as comments from residents about its effectiveness.

P46. The limbic system in the light of modern neurobiology: basic and clinical implicationsVassilis E. Koliatsos, Jun Yan, Myriam Gastard, Shelley Weaver, Susan Bora (Departments of Pathology [Neuropathology], Neurology, Neuroscience, and Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, MD)

The concept of the limbic system has had vast implications for contemporary psychiatry and behavioral neurology. Introduced by Broca to signify an across-species stability of a portion of the mesial hemisphere, the limbic system was declared by Papez to be the site of obligatory emotional processing and was demonstrated by MacLean to be critical in the elaboration of mammal-specific behaviors. Although part of the lexicon of modern neuroscience, this ambiguous concept has remained notoriously unpopular among basic scientists. Recent work in our laboratory lends new credence to the uniqueness of the hippocampus/olfactory cortex and directly connected areas by showing that a) neurogenesis is a central theme in all these structures and probably extends beyond the originally observed “hotbeds” in the dentate gyrus and olfactory bulb under conditions of injury and b) pyramidal neurons in limbic and paralimbic areas are especially vulnerable to apoptosis, during injury, via the activation of a dormant, apoptosis-resistant system of nitric oxide–releasing interneurons; this mechanism is relevant to the early massive death of entorhinal neurons in Alzheimer's disease. We argue that the potential for neurogenesis and vulnerability to cell death lead to ongoing restructuring of core limbic structures under stress or injury and may play significant roles in the establishment and malleability of autobiographical memory.

Traumatic Brain Injury

P47. Low-dose donepezil normalizes P50 physiology in traumatic brain injury patientsDavid B. Arciniegas, Jeannie L. Topkoff, C. Alan Anderson, Christopher M. Filley, Lawrence E. Adler (Denver Veterans Affairs Medical Center, Denver, CO)

Background: Traumatic brain injury may produce hippocampal cholinergic dysfunction. Cholinergic dysfunction may be reflected by abnormal suppression of the P50 evoked waveform to paired auditory stimuli (P50 ratio). We previously demonstrated normalization of P50 physiology with donepezil HCl 5 mg in an uncontrolled case series (Arciniegas et al, J Neuropsychiatry Clin Neurosci 2001; 13:140). The present randomized double-blind placebo-controlled crossover design study was performed to determine whether cholinergic augmentation would normalize P50 physiology. Methods: 10 subjects with remote TBI of at least mild severity (American Congress of Rehabilitation Medicine criteria; J Head Trauma Rehabil 1993; 8:86–87), persistent symptoms of impaired auditory gating, and abnormally large P50 ratios as determined by our previously described methods (Arciniegas et al, J Neuropsychiatry Clin Neurosci 2000; 12:77–85) were recruited for the study. All were naive to procholinergic treatment. Subjects were treated with donepezil HCl 5 mg daily or placebo for 6 weeks, followed by donepezil HCl 10 mg daily or placebo for another 6 weeks; at week 13, subjects were crossed over to either donepezil or placebo and repeated this medication schedule. Results: There was a significant effect of treatment condition on P50 ratio (F=4.7, df=4,36, P<0.004). There was a significant reduction in P50 ratio on donepezil HCl 5 mg daily when compared with baseline (P=0.04) and with each of the other treatment conditions (all P≤0.05). P50 ratio on donepezil HCl 10 mg daily and both placebo conditions did not differ significantly from baseline. Conclusions: Low-dose donepezil normalized P50 physiology, offering support for the hypothesis that cholinergic dysfunction may contribute to the development of abnormal P50 physiology and related symptoms following TBI. The observation that high-dose donepezil was associated with reemergence of abnormal P50 physiology suggests that there may be a therapeutic window for treatment response in this population, outside of which therapeutic benefits may be reduced.Support from Department of Veterans Affairs and Pfizer, Inc.

P48. Evaluation of cognitive impairment and functional outcome of brain-injured patients using the Repeatable Battery for the Assessment of Neuropsychological StatusGeorge Michael Cuesta, Alan Blau, Karina Ortega-Verdejo, Barry Jordan (Burke Rehabilitation Hospital, White Plains, NY)

Background: The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) has been found to predict functional outcome in brain-injured patients as measured by the Functional Independence Measure (FIM). However, it has not been correlated with the Galveston Orientation and Amnesia Test (GOAT), the Mini-Mental State Examination (MMSE), or length of stay (LOS). Methods: Cognitive function of 87 patients on an acute rehabilitation inpatient brain injury unit was evaluated by using the RBANS, the GOAT, and the MMSE over an 18-month period. Outcome was measured with the FIM and LOS. Results: Total RBANS index scores correlated significantly with the total GOAT (r=0.336, P=0.007) and MMSE (r=0.565, P=0.000) scores. Total admission and discharge FIM scores (r=0.420, P=0.000, and r=0.317, P=0.003, respectively) significantly correlated with total RBANS index scores. Length of stay demonstrated an inverse relationship with total RBANS scores (r=–0.325, P=0.002). The average total RBANS index score for the sample was 71. The most impaired subtest scores were attention (72) and delayed recall (72) indexes. Visuospatial/constructional index was least impaired (84). Conclusions: The significant relationship between performance on the RBANS, GOAT, and MMSE suggests that these instruments measure similar neuropsychological constructs. Our study confirms earlier findings of a significant relationship between RBANS performance and total FIM scores and suggests that length of stay may be another significant outcome variable.

P49. Use of the Cognitive Test for Delirium in TBI patientsRichard E. Kennedy, Risa N. Thompson, Todd G. Nick, Mark Sherer (Medical College of Virginia, Richmond, VA; Mississippi Methodist Rehabilitation Center, Jackson, MS; University of Mississippi Medical Center, Jackson, MS)

Background: Delirium is a common consequence of TBI, although exact prevalence is unknown. In part, this uncertainty arises from difficulty in assessing the impaired TBI patient. The Cognitive Test for Delirium (CTD; Hart et al, Psychosomatics 1996; 37:533–546) was developed to diagnose delirium in the intensive care unit with patients having verbal and motoric impairments. We hypothesized the CTD would be useful in TBI patients with similar deficits. Methods: Subjects were 65 consecutive patients at Rancho Los Amigos scale IV or above in the TBI Model System. Each patient was evaluated weekly by a trained clinical neuropsychologist who determined whether the patient met DSM-IV criteria for delirium. An independent research assistant administered the CTD to each patient. Results: 45 patients initially met criteria for delirium, which resolved in 31 during hospitalization. Complete ratings were available for 249 of 304 weekly observations. A receiver-operating characteristic (ROC) curve suggested an optimum cutoff value of 22 for the CTD, with an area under the curve (AUC) of 0.7978. At this value, sensitivity was 72% and specificity was 71%. Conclusions: The CTD differentiates between delirious and nondelirious TBI patients at acceptable levels. Further studies of its usefulness compared with other tests are warranted.Support from NIDRR Grant H133A9980035 to TBI Model Systems of Mississippi.

P50. Role of COMT Val 108/158 Met genotype in executive functioning following TBIRobert H. Lipsky, Molly B. Sparling, Laurie M. Ryan, Ke Xu, Andres M. Salazar, David Goldman, Deborah L. Warden* (Defense and Veterans Head Injury Program, Walter Reed Army Medical Center, Washington, DC; Laboratory of Neurogenetics, National Institute on Alcohol Abuse and Alcoholism, NIH, Bethesda, MD)

Background: In the central nervous system, the enzyme catechol-O-methyltransferase (COMT) inactivates neurotransmitters such as dopamine and norepinephrine. Dopamine is linked to many cognitive abilities, including executive functioning. High enzyme activity (COMT Val) and low enzyme activity (COMT Met) variants are functional polymorphisms that influence dopamine levels in the brain. To examine the role of a functional COMT polymorphism in executive functioning following traumatic brain injury, 123 individuals were evaluated. Methods: Subjects underwent a full evaluation including neuropsychological and genetic testing. COMT Val/Met genotype determinations were performed by using a 5′ nuclease assay. Results: Of the total sample, 43 were homozygous for COMT Val, 27 were homozygous for COMT Met, and 50 individuals were heterozygotes. ANOVA of perseverative responses on the Wisconsin Card Sorting Test (WSCT) was significantly different between genotype groups (P=0.01). As observed in other populations, individuals homozygous for the COMT Met allele made fewer perseverative responses on the WCST (mean=11.78), whereas those homozygous for COMT Val had the highest number of perseverative responses (mean=20.47). Heterozygotes made an intermediate number of responses (mean=13.23). Conclusions: Our data support the influence of COMT Val 108/158 Met in cognitive function, extending the relationship to subjects in whom executive function has been impaired by injury.

P51. Working memory activation patterns one month and one year after mild traumatic brain injury: a longitudinal fMRI studyThomas W. McAllister, Molly B. Sparling, Laura A. Flashman, Brenna McDonald, Heather Wishart, Andrew J. Saykin (Dartmouth Medical School, Lebanon, NH; New Hampshire Hospital, Concord, NH)

Background: We have previously reported differences in brain activation patterns associated with working memory (WM) tasks in patients studied 1 month after mild traumatic brain injury, despite equivalent WM performance. We hypothesized that individuals with MTBI would be less symptomatic and would show normalization of their activation patterns 1 year after their injuries. Methods: 12 MTBI patients (Glasgow Coma Scale score 13–15, loss of consciousness <30 minutes) were studied 1 month and 1 year after their injury and compared with 6 healthy control subjects also studied twice, 1 year apart. Both groups completed symptom checklists, neuropsychological assessment, structural MRI, and whole-brain echo-planar functional MR imaging (fMRI; 1.5 T). Participants performed an auditory N-Back task with variable load requirements during fMRI. Results: The MTBI group had significantly more cognitive complaints and cognitive deficits (primarily speed of information processing) 1 month after injury. One year after injury, the MTBI group had fewer cognitive complaints and deficits. fMRI activation in regions associated with working memory occurred in both groups, with increased activation in response to higher processing demand as previously reported. There were significant time by diagnosis by activation interactions associated with processing load (0,1,2,3–back) in areas of WM circuitry. Conclusions: The observed changes in fMRI activation patterns between groups over time suggest that changes in the activation and modulation of WM processing resources may underlie the subjective and objective cognitive improvement seen in most individuals who sustain MTBI.Support from NIDRR Grants H133G70031 and H133G000136, and from NIH Grant RO1 NS40472-01.

P52. An examination of the high dropout rate in studies of patients with mild traumatic brain injuryScott McCullagh, Katy Kamkar, Alison Jardine, Andrea Protzner, Nancy Blair, Anthony Feinstein (University of Toronto, Toronto, ON, Canada).

Background: High dropout rates in mild TBI research raise questions about the external validity of study results. In Canada, this issue can be examined by using provincial health insurance databases to obtain information about lost subjects, even without their participation in a study, that would be otherwise unavailable. Methods: 626 consecutive patients with mild TBI were asked to participate in TBI outcome research. Those who accepted (n=272) were compared with those who refused or dropped out after enrollment (n=354) on demographic and injury-related variables. Thereafter, using encrypted health card data, the two groups were contrasted with respect to pre- and post-TBI health care utilization. Results: No demographic differences between the groups emerged. However, TBI indices were different, with longer duration of posttraumatic amnesia and higher rates of CT abnormalities for those who participated in studies (P<0.001). Similarly, healthcare utilization rates were no different prior to injury, but were significantly elevated postinjury for study participants, even after 6 months (P<0.001). Conclusions: Demographic and premorbid variables did not predict whether patients comply with or refuse study participation. However, the study sample was biased toward those with more significant injuries, which translated into greater health care utilization. These results strike a cautionary note for mild TBI research.Support from Ontario Trauma Foundation and Physicians Services Incorporated.

P53. Natural history of cognitive, affective, and physical symptoms of postconcussion syndrome following mild traumatic brain injuryTara McHugh, Robert Laforce, Patricia Gallagher, Susan Quinn, Phillip Diggle (University of New Brunswick, Saint John, NB, Canada)

Background: Mild traumatic brain injury is associated with cognitive, affective, and physical changes (Kay et al, J Head Trauma Rehabil 1993; 8:86–87). When these symptoms persist for more than 3 months, a diagnosis of postconcussion syndrome (PCS) is often given (American Psychiatric Association, DSM-IV, 1994). The aim of this study was to track PCS development in cognitive, affective, and physical realms in an attempt to better explore the factors leading to its evolution from MTBI. Methods: Out of 173 consecutive admissions at a large regional facility, 93 fit our inclusion criteria, and 26 MTBI patients agreed to receive a comprehensive neuropsychological assessment, at three intervals: within 1 month, at 3 months, and at 6 months post concussion. Results: On the basis of DSM-IV criteria and clinical judgment, 2 external raters diagnosed 4 MTBI patients with PCS (15.4%). Results suggested that the neuropsychological profile of the PCS patients was significantly different from our age-, gender-, education- and IQ-matched control subjects and MTBI patients who did not develop PCS (84.6%). Interestingly, all PCS patients seemed to improve in their overall level of functioning from 3 to 6 months, but they never showed similar levels of functioning at 6 months to MTBI patients who did not develop PCS. Conclusions: Although based on a small sample of participants, these results provide further insight into the active ingredients of the natural history of PCS.Support from the Rick Hansen Neurotrauma Initiative (Medical Research Council of Canada).

P54. The clinical significance of major depression following mild traumatic brain injuryMark J. Rapoport, Scott McCullagh, Anthony Feinstein (University of Toronto, Toronto, ON, Canada)

Background: Major depression (MD) is a common sequela to mild traumatic brain injury. To what extent its presence may affect outcome has not, however, been ascertained. Methods: A consecutive sample of 184 patients meeting the American Academy of Rehabilitation Medicine criteria for MTBI were assessed for MD within a mean of 45.71 days (SD=29.7) of injury with the Structured Clinical Interview for DSM-IV disorders. Those with MD (30/184, 16.3%) were compared to those without (154/184, 83.7%) on measures of psychosocial outcome (Rivermead Follow-up Questionnaire; RFUQ), psychological distress (General Health Questionnaire; GHQ), symptoms of postconcussion disorder (Rivermead Postconcussion Disorder Questionnaire; RPDQ) and Global Outcome (Glasgow Outcome Scale; GOS). Results: Patients with MD were more likely to be female (P<0.05), to be younger (P<0.05), and to misuse substances (P<0.05). Patients with MD endorsed significantly more difficulties on the RFUQ, GHQ, and RPDQ (P<0.0001), even controlling for age, gender, and substances. There were no differences on the GOS. Conclusions: The presence of MD during the acute phase following mild TBI is associated with more psychosocial difficulties as well as more prominent postconcussion symptoms. Our study highlights the need for the early diagnosis and prompt treatment of major depression following mild TBI.

P55. Cognitive and personality tests as predictors of employability after brain injuryDavid Salisbury, Julie L. Williams, Jeffery Allen, William Arnold (Wright State University, Dayton, OH)

Background: Research has shown that personality variables and cognitive variables are stronger predictors of employment for persons with brain injury when used together (Heaton et al, J Nerv Ment Dis 1978; 166:408–416). The intent of this research proposal is to determine which combinations of cognitive and personality variables are most predictive of vocational outcome. Methods: Archival data are being collected from an outpatient rehabilitation program. Data will be gathered from 75 to 100 subjects with varying degrees of brain injury sustained between the years 1995 and 2000. Results: A preliminary multiple stepwise regression was performed on 42 subjects (32 men and 10 women) ages 21 to 61. Personality variables were taken from the MMPI-2. Cognitive variables included indices from the Wechsler Adult Intelligence Scale–III and the Wechsler Memory Scale–III. The Auditory Delayed Recall Scale (F=7.78841, df=2,21, P=0.0304) and the Psychasthenia Scale (F=7.78841, df=2,21, P=0.0394) were found to be a significant predictors of wage. Conclusions: Personality and cognitive variables continue to demonstrate predictive value in regard to employment status after brain injury.

P56. Positive influence of acute alcohol use on outcome following traumatic brain injuryMolly B. Sparling, Laurie M. Ryan, Joan Walter, Deborah L. Warden (Defense and Veterans Head Injury Program, Walter Reed Army Medical Center, Washington, DC)

Background: Intoxication leads to a greater risk of traumatic brain injury due to impaired judgment and increased risk-taking behavior. However, studies examining the effect of alcohol at the time of injury on recovery from TBI have been contradictory. Measures of outcome, alcohol at injury, prior history of alcohol abuse, and postinjury drinking habits were analyzed to examine the effects of acute pre- and postinjury alcohol use in a military TBI population. Methods: A group of 148 active-duty service members who sustained a TBI participated in a comprehensive evaluation and rehabilitation protocol. Individuals provided informed consent and protocols were reviewed and approved by the Walter Reed Institutional Review Board. Alcohol use at the time of injury was analyzed in relation to measures of outcome, prior history of alcohol abuse, and postinjury drinking habits. Results: Data on alcohol use at the time of injury were available on 138 of 148 individuals in the study. Of the 138, injury was related to alcohol use in 39 individuals. There were no differences between groups on demographic variables including measures of premorbid functioning, age, education, handedness, race, or severity of injury. Individuals using alcohol at injury had significantly greater past history of substance abuse (P≤0.0001) and continued drinking after their injuries (P≤0.0001). Despite this, they had better Glasgow Outcome Scores at 7 days post injury (P≤0.05). Additionally, the alcohol at injury group had significantly higher rates of return to duty at 1 year post injury (P≤0.05). Prior history of substance abuse did not predict return to duty. Conclusions: In this population, use of alcohol at the time of injury appeared to have a neuroprotective effect that led to better initial and longer-term recovery.

P57. Clinical correlates of aggressive behavior after traumatic brain injuryAmane Tateno, Ricardo Jorge, Robert G. Robinson (University of Iowa, Iowa City, IA)

Background: Aggressive behavior is one of the frequent and severe problems occurring after traumatic brain injury. In the present study we assessed aggression using a quantitative scale and examined its clinical correlates. Methods: We studied 89 TBI patients and 26 patients with multiple traumatisms but without neurological involvement; the latter served as a control group. Aggressive behavior was assessed by using the Overt Aggression Scale (OAS). Psychiatric diagnosis was done by using a semistructured interview. Severity of depressive and anxiety symptoms was assessed by using Ham-D and Ham-A. Impairment variables included the MMSE, Functional Independence Measure, Social Functioning Exam, and Social Tie Checklist. Results: Of the 89 TBI patients, 30 showed significant aggressive behavior after resolution of posttraumatic amnesia and during the first 6 months after TBI. A diagnosis of major depression was significantly more frequent in the aggressive group. In addition, aggression was significantly associated with the presence of frontal lobe lesions, poor premorbid social functioning, and a history of alcohol and substance abuse. Conclusions: Aggressive behavior after TBI is related to the presence of major depression, substance abuse, poor social functioning, and damage to frontal lobe.Support from NIMH Grants MH40355, MH52879, MH53592, and Research Award MH00163, and a grant from the Department of Neuropsychiatry, Nippon Medical School, Tokyo, Japan.

P58. Comparison of cognitive impairment associated with major depression following stroke versus following traumatic brain injuryAmane Tateno, Yuichi Murata, Robert G. Robinson (University of Iowa, Iowa City, IA)

Background: The association of major depression with cognitive impairment has been found in stroke patients, but not in patients with traumatic brain injury. This study tested the hypothesis that age differences would account for the lack of effect of major depression on cognitive function following TBI. Methods: We examined cognitive function of 354 stroke patients and 154 TBI patients using the Mini-Mental State Examination. 73 of 354 stroke patients and 35 of 154 TBI patients received diagnoses of major depression based on the Present State Examination, a semistructured interview. Background and neurological characteristics as well as Ham-D and neuroimaging of the lesions were obtained. Results: Stroke patients with major depression were significantly older, less educated, more cognitively impaired, and had significantly larger lesion volumes than TBI patients with major depression. Stroke and TBI patients were matched for years of education and lesion volumes with no change in the findings related to cognitive dysfunction. When patients were matched for age (±3 years), there was no longer any association of major depression with cognitive impairment. Conclusions: The association between major depression and cognitive impairment is present only in elderly stroke patients. Our results support the hypothesis that dementia of depression in both brain-injured and non–brain-injured patients occurs only in the elderly.Support from NIMH Grants MH40355, MH52879, MH53592, and Research Award MH00163, and a grant from the Department of Neuropsychiatry, Nippon Medical School, Tokyo, Japan.

Imaging

P59. Mapping and quantification of dopamine D₂ receptor activationKevin J. Black, Tamara Hershey, Jonathan M. Koller, Juanita L. Carl, Joel S. Perlmutter (Washington University School of Medicine, St. Louis, MO)

Background: Neuronal circuits influenced by dopamine D₂-like receptors mediate the effects of neuroleptic drugs and may be relevant to the pathophysiology of several human disease states, such as schizophrenia, drug abuse, akathisia, or Tourette syndrome. We tested the validity of a D₂ “challenge test.” Methods: We used PET and arterial radioactivity sampling in 5 normal baboons to measure the regional cerebral blood flow (rCBF) responses produced by an acute dose of the dopamine D₂-preferring agonist U91356a (1–200 μg/kg, iv). Images were referred to the B2K baboon template (http://purl.org/net/kbmd/b2k) and analyzed, using SPM99 to correct for multiple comparisons. A partial analysis was previously reported (J Neurosci 1997; 17:3168–3177). Results: Absolute average brain blood flow decreased in a dose-responsive fashion (–30% at the highest dose). In some regions, flow decreased significantly more than this whole-brain average (peak t-score>7.5, df=34). These regions include dorsolateral prefrontal cortex, anterior putamen, medial parietal cortex, primary sensorimotor cortex, posterior cingulate, and thalamus. Conclusions: These results lay the groundwork for interpretation of future similar studies in disease models or in humans.Support from NINDS, Dana Foundation, NARSAD, Tourette Syndrome Association, American Parkinson Disease Association, and McDonnell Center for Higher Brain Function. U91356a was supplied by Pharmacia-Upjohn.

P60. Evaluating the contributions of amygdala and other brain volumes to individual differences in personality/temperament traitsPaul T. Costa, Jr., Godfrey Pearlson (NIA/NIH, Bethesda, MD; Johns Hopkins University School of Medicine, Baltimore, MD)

Background: Neurobiological research on emotional brain circuits consistently implicates the amygdala as a central structure, especially in preclinical research on fear mechanisms and fear conditioning. The amygdala may also participate in positive emotional behaviors as well. Depue and Collins (Behav Brain Sci 1999; 22:491–569) cite the amygdala's role in integrating incentive motivation through the projections of the so-called extended amygdala, leading some to consider it as an emotional motor system. Well-documented pancultural differences by sex in personality and temperamental traits suggest that the sex of the individual modulates contribution of the amygdala and related subcortical structures to the phenotypic variation in personality traits. Methods: For 197 male and female participants in the Aging and Brain Cognition study, brain MRI scans and NEO Five-Factor Inventory scores from the Five-Factor Model (FFM) of personality were examined to determine differential associations between various brain volumes and personality dimensions. Results: Raw Pearson correlations showed significant associations for left, right, and total amygdala volumes for women but not men. Results were unchanged after partialing total brain volume. Right and total amygdala volumes were significantly associated with Openness to Experience (O) scores for men but not women. Conclusions: A series of hierarchical regression analyses confirmed that amygdala as well as putamen and caudate volumes make independent contributions to explaining individual variation in FFM personality dimensions.Support from NIH Grant R01 AG11859 and the NIA Intramural Research Program.

P61. How does high-resolution brain SPECT help in ADHD?R. Kohn, D.G. Pavel, G. Davis, P. Epstein, R. Grant, P. Kim, I. Craita, Z. Taha (University of Illinois Medical Center, Chicago, IL)

Background: An increasing body of evidence points to the usefulness of high-resolution brain SPECT in attention-deficit/hyperactivity disorder. Nevertheless, a tendency to downplay its use persists. One of the objective difficulties is that attention problems are often complicated by behavioral and mood disorders, which in themselves can also affect attention. Methods: 243 patients were referred for attention disorders. All of them had additional symptoms suggesting one or several combined conditions. Tc-HMPAO was used in conjunction with a triple-head 3000 Prizm gamma camera, and four types of color displays were generated. Results: Distinct visualization was possible of various areas involved in attention circuits: prefrontal areas, orbitofrontal cortices, anterior cingulate, striatum, and various portions of the thalamus, as well as paralimbic structures. In addition, the perfusion pattern of the cortex was identified. Conclusions: Given an appropriate technique, SPECT can reliably identify cortical and subcortical structures as hyper- or hypoperfused. Careful knowledge of the perfusion pattern in these structures involved in the frontal-subcortical circuits has led to the expansion of our treatment options beyond a clinical symptomatic approach.

P62. Age-associated changes in substantia nigra microstructural integrity: a diffusion tensor imaging studyDavid A. Medina, Maria C. Carrillo, Michael E. Moseley, Glenn T. Stebbins (University of Illinois at Chicago/Rush-Presbyterian-St. Luke's Medical Center, Chicago, IL)

Background: Age-associated degeneration of dopaminergic cells in the substantia nigra (SN) has been documented (Gabrieli, Proc Natl Acad Sci USA 1996; 93:13534–13540). Declines in dopaminergic function are associated with decreased motor and cognitive abilities (Bennett et al, N Engl J Med 1996; 334:71–76). In vivo documentation of SN integrity has been difficult to achieve. Diffusion tensor imaging (DTI) is a new development of MRI technology that provides a measure of microstructural integrity of the neural tissue. This study compared SN integrity in healthy young and old persons by using DTI. Methods: 20 right-handed healthy subjects participated; 10 younger (mean=29 years) and 10 older (mean=80 years). DTI data were processed to provide a measure of integrity: fractional anisotropy (FA). Midbrain regions of interest were prescribed and compared for all subjects. Results: Midbrain FA was significantly reduced in healthy aging (P<0.05 controlled for multiple comparisons). Two regions showed these differences: 1) transverse pontine fibers and cerebral peduncle and 2) bilateral SN. Conclusions: Decreases in the microstructural integrity of SN are present in healthy older persons. Age-associated decrease in SN integrity appears to document age-associated changes in dopaminergic function. These results may provide an explanation for the motor and cognitive declines present in normal aging.

P63. While achieving normal learning, patients with mild Parkinson's disease exhibit widespread dysfunction within the normal sequence-learning brain networkMarc J. Mentis, Dennis Zgaljardic, Dominique Delalot, Vijay Dhawan, Claude Ghez, David Eidelberg (Center for Neurosciences, North Shore University Hospital, Manhasset, NY)

Background: Because the trial-and-error motor sequence learning task (TEseq) has a high working memory load, it is well suited to evaluate the “frontal” cognitive dysfunction observed in nondemented Parkinson's disease patients. Methods: Learning during TEseq was constrained so as to be equivalent between 8 PD patients and 8 age-matched control subjects by the end of a PET scan. Voxel-based t-tests were evaluated by using statistical parametric mapping, while differences in network activation, and brain–behavior correlations were analyzed using partial least squares. Results: TEseq activated 4 times as much brain in the PD group compared with the control subjects. The regions activated by the PD group included 1) those regions activated by control group: right dorsolateral prefrontal cortex, right premotor cortex, right parietal cortex, rostral supplementary area, and precuneus, 2) regions not activated by the control group: the left-sided homologs of the above regions and bilateral lateral cerebellum. Task salience values were 15% greater in the PD group, and correlations between sequence learning and the brain network were significantly different between groups. Conclusions: To achieve equal learning, the PD patients utilized more of the normal sequence-learning network with about 15% less efficiency than the control subjects. Inefficiency appeared to be caused by differences in how nodes within the network related to behavior.Support from NIH Grant K23 NS02204-02.

P64. Good-quality brain SPECTS: necessary and sufficient prerequisitesD.G. Pavel, R. Kohn, G. Davis, P. Epstein, B. Chuang, T. Liu, Z. Taha, I. Craita (University of Illinois Medical Center, Chicago, IL)

Background: Brain SPECT (relative perfusion mapping), has not yet achieved wide recognition in psychiatry. Common comments by potential referring physicians are that it is an inadequate modality and does not influence patient management. Indeed, poor-quality images are often provided. We implemented a number of necessary features for quality images and tried to establish if they are sufficient. Methods: Multiple displays were used for the past 3 years, including standard orthogonal sections, special temporal sections, self-explanatory color codes with discrete shades, surface images, 3D multiple threshold volume images, automatic realignment of images in cases of follow-up, and statistical parametric mapping (SPM) evaluation of the significance or change. Results: With the exception of SPM, this constellation of display methods has enabled a detailed evaluation of cortical as well as subcortical structures by the specialist and provided clear images for the referring physician. The semiquantitative measurement by special color tables is fast and useful. Quantification by SPM, while promising, is very dependent on a high-quality, age-matched normal database. Conclusions: Necessary technical features have been established for visualization of image details and for semiquantification. These are sufficient until a high-quality normative database enables SPM analysis.

P65. Morphology of the ventral frontal cortex in adult males with nonsyndromic clefts of the lip and/or palate: relationship to social functionPeg C. Nopoulos, Stephanie Berg, John Canady, Duane Van der Mark, Lynn Richman, Nancy Andreasen (University of Iowa Hospitals and Clinics, Iowa City, IA)

Background: In a previous MRI study, adult males with nonsyndromic clefts of the lip and/or palate (NSCLP) were found to have increased volumes of frontal lobe gray matter. Moreover, many subjects with NSCLP have been described as being socially inhibited. The ventral frontal cortex (VFC) has been shown to be related to social function. This study was designed to evaluate the morphology of the VFC in adult males with NSCLP and its relationship to social function. Methods: Subjects were 43 males with NSCLP and 43 sex-matched control subjects. Social function was assessed by using a standardized scale. The morphology of the VFC was obtained from MRI scans using the software BRAINS. Results: After controlling for frontal lobe gray matter, the patient group had significantly smaller VFC volumes on both the left (F=4.08, P=0.04) and the right side (F=5.28, P=0.02). The volume of the VFC was significantly and inversely correlated to a measure of social function (r=–0.37, P=0.02). Conclusions: Despite significant excess of overall frontal lobe cortical volume in subjects with NSCLP, the VFC was abnormally small in volume compared with matched control subjects. Moreover, the structural abnormality in this brain region was directly correlated with abnormalities in social function.Support from K-23 Grant DE00417-01 from the NIH National Institute of Dental and Craniofacial Research.

P66. Endocrine and neural correlates of induced sadness and elationWilliam Ottowitz, Darin Dougherty, Robert Connors, Walter Brown (Massachusetts General Hospital, Psychiatric Neuroscience Program, Charlestown, MA)

Background: Several functional imaging studies have investigated the neural correlates of normal affective states. Projections from the neuroanatomical substrates of these emotional states to the hypothalamic-pituitary-adrenal axis warrants simultaneous investigation of concurrent hormonal responses. Defining several of the neurobiological correlates of normal emotional states has ramifications for the differentiation of normal from pathological affective states. Methods: Six nonpsychiatric female subjects received Velten sadness, elation, and neutral mood inductions with subsequent evaluation by SPECT. This was coordinated with acquisition of cortisol levels and with assessment of mood by the visual analog scale. Results: Paired t-test comparison of the scans acquired during sadness induction demonstrated decreased activity of the left thalamus, left dorsolateral, parietal, and medial frontal cortices relative to the neutral inductions, whereas elation induction demonstrated decreased activity of the medial frontal cortex bilaterally and of the left basal ganglia. Cortisol levels demonstrated significant increases during inductions of sadness, but not elation. Conclusions: Affective dysregulation is central to several neuropsychiatric disorders, including, e.g., mood disorders and pseudobulbar palsy. Our findings for the neural and endocrine correlates of sadness parallel those of major depression, but the functional imaging studies of mania and pseudobulbar palsy are currently too few to offer an adequate comparison to studies of induced elation. These comparisons deserve further investigation and may generate an objective means to distinguish pathological from normal affective states.Support from Innovative Medical Imaging, Waltham, MA.

P67. Association of caudate and putamen volumes with Five-Factor Model dimensions of personality (NEO FFI scores)Godfrey Pearlson, Paul T. Costa, Jr. (Johns Hopkins University, Baltimore, MD; NIMH, Rockville, MD)

Background: Left caudate and left putamen areas have high concentrations of dopamine, and Fisher et al. (Pers Individ Dif 1997; 23:345–352) have shown greater relative rCBF in these same regions. We hypothesized that caudate and putamen volumes would be related to certain personality dimensions related to approach motivation, activity, and exploration. In terms of the dimensions of the Five-Factor Model of personality (NEO Five-Factor Inventory; NEO FFI), we would expect associations of caudate and putamen volumes with Extraversion (E) and Openness to Experience (O). Methods: Data gathered from 193 participants in the Aging and Brain Cognition cohort were used in hierarchical multiple regression models. Brain volumetric measures were regressed on the NEO FFI dimensions to examine the contribution of cortical and subcortical structures to Neuroticism (N) and Extraversion as well as the other dimensions of the model. Results and Conclusions: After ventricle-to-brain ratio and sex were entered in separate steps, left and right volume scores for frontal lobes, hippocampus, amygdala, caudate, and putamen were entered. For E scores, putamen, caudate, and frontal lobe volumes were the significant predictors, supporting the hypothesis that brain areas rich in dopamine receptors or neurons would be associated with individual differences in approach and exploratory traits.Support from NIH Grant R01 AG11859 and the NIA Intramural Research Program.

ABBREVIATIONS

The abbreviations defined below are used frequently within the abstracts.

AD=Alzheimer's disease; ADHD=attention-deficit/hyperactivity disorder; ANCOVA=analysis of covariance; ANOVA= analysis of variance; CT=computed tomography; ECT=electroconvulsive therapy; EEG=electroencephalography; FSIQ=Full Scale IQ; Ham-A=Hamilton Rating Scale for Anxiety; Ham-D=Hamilton Rating Scale for Depression; HD=Huntington's disease; MANOVA=multivariate analysis of variance; MMPI=Minnesota Multiphasic Personality Inventory; MMSE=Mini-Mental State Examination; MR=magnetic resonance; MRI=magnetic resonance imaging; MS=multiple sclerosis; MTBI=mild traumatic brain injury; NARSAD=National Alliance for Research on Schizophrenia and Depression; NIA=National Institute on Aging; NIMH=National Institute of Mental Health; NIH=National Institutes of Health; NINDS=National Institutes of Neurological Disorders and Stroke; OCD=obsessive-compulsive disorder; PD=Parkinson's disease; PET=positron emission tomography; PIQ=Performance IQ; rCBF=regional cerebral blood flow; SD=standard deviation; SPECT=single-photon emission computed tomography; SSRI=selective serotonin reuptake inhibitor; TBI=traumatic brain injury; VIQ=Verbal IQ.

*Corresponding author (if other than the author listed first).