Diagnosis and Neuroanatomical Correlates of Depression in Brain-Damaged Patients Implications for a Neurology of Depression Elliott D. Ross, MD, A. John Rush, MD \s=b\ Recognizing depression in brain-damaged patients poses considerable problems. The standard dignostic criteria often are not applicable since the neurological lesion may distort or even obliterate salient features of depression. Patients actually may deny being depressed or dysphoric, not have a depressive affect, or be totally unaware of abnormal vegetative behaviors. Furthermore, brain lesions themselves may produce striking behavioral alterations that can be mistakenly attributed to depression, or striking behavioral changes due to depression may be mistakenly attributed to the brain lesion. Based on five case studies, several clinical guidelines for recognizing and diagnosing depression in brain-damaged patients are offered. These cases also provide a data base to generate initial hypotheses about the neuroanatomical basis of the depressive syndrome. By observing how focal brain lesions modify the signs and symptoms of depression, inferences are made about brain areas crucial for modulating the various features of the depressive syndrome. (Arch Gen Psychiatry 1981;38:1344-1354) major problem confronting neurology define chiatry A psychiatric discovery antidepressants, both and psy¬ is to the neuroanatomical basis of of disorders. The anxiolytics, antipsychotics, and neuroendocrine markers strongly argues that specific biological alterations in brain function occur in certain psychiatric disorders but contrib¬ utes little to understanding their neuroanatomical basis. Changes in the normal patterns of interhemispheric and intrahemispheric processing during motor, sensory, lan¬ guage, and cognitive activities have been observed in both affective and schizophrenic disorders with the use of clinical, neuropsychological, linguistic, and neurophysiological techniques (eg, spectral analysis of EEGs, internal carotid amobarbital sodium [Amytal] injections, unilateral electroconvulsive therapy [ECT], directional preponder¬ ance of nystagmus during caloric stimulation, eye move¬ ments during mental tasks, electrodermal changes during the orienting response, testing of semantic and syntactical processing of language, focal cerebral blood flow determi¬ tests, etc1'3). From these data, nations, dichotic listening Flor-Henry and Koles' ' have suggested that depression, dysphoria, anxiety, and sadness relate to functional alter¬ ations in the frontotemporal regions of the right hemi¬ sphere, while mania, paranoia, euphoria, and anger relate to functional alterations in the left frontotemporal regions. Nevertheless, the precise anatomical correlates for the cognitive and behavioral abnormalities encountered in various psychiatric disorders remains unclear. Accepted for publication June 24, 1981. From the Departments of Neurology (Dr Ross) and Psychiatry (Drs Ross and Rush), University of Texas Health Sciences Center at Dallas. Reprint requests to Department of Neurology, University of Texas Health Science Center, 5323 Harry Hines Blvd, Dallas, TX 75235 (Dr Ross). Recent neurological investigations by Heilman et al,14 Tucker et al,'5 and Ross and co-workers'6'8 have suggested a distinct and dominant functional-anatomical organiza¬ tion in the right hemisphere for the modulation of the affective components of language and behavior. These investigations complement many psychological, physiologi¬ cal, and behavioral studies that previously have implicated the right hemisphere in the modulation of emotions,4''2'0"24 a notion first hypothesized by Jackson25 nearly a century ago. The disorders of affective language and behavior resulting from focal lesions of the right hemisphere are called "aprosodias."" The aprosodias can be classified" in a manner similar to the aphasias,26 which are disorders of propositional language (syntax and word choice) that are usually the result of focal lesions in the left hemisphere (Table 1). These findings strongly imply that the signs and symptoms encountered in some of the major affective disorders also might have a neuroanatomical basis. As with much of our accrued knowledge about the organization of the human brain, both the aphasias and aprosodias were discovered and defined by using the classic neurological method of inferring function through clinicopathological correlations. Similarly, the neuroana¬ tomical basis for psychiatric disorders might be inferred from studying patients who have both a psychiatric disor¬ der and a neurological lesion. If a particular brain lesion reduced or obliterated one or more of the signs and symptoms associated with the psychiatric disorder, then one could infer that the lesioned portion of the brain was critical in modulating that particular sign or symptom. In attempting to carry out this strategy in patients with depression and brain damage, we rapidly discovered that our major impediment was to diagnose reliably the depres¬ sive syndrome. The remarkably consistent descriptions of depression from the time of Hippocrates27 have allowed the develop¬ ment of specific criteria, such as the Research Diagnostic Criteria,2" Washington University criteria,20 and DSMIII,30 for diagnosing depression in patients without brain damage. In brain-damaged patients, however, we found that the neurological lesion may either obliterate or distort criterion signs and symptoms or may prevent collection of adequate historical information because of propositional and affective language disturbances. Furthermore, the depressive disorder can modify the clinical course usually associated with a neurological lesion, and changes in affective behavior resulting from a brain lesion may be misconstrued as being etiologically related to a depressive syndrome. Thus, the usual diagnostic criteria for depres¬ sion may be inapplicable in these clinical situations. The problems of disentangling the signs and symptoms of depression from those caused by brain damage are exemplified by the study of Folstein et al.3' Although they found an increased incidence of selected depressive fea- Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 Table 1.—The Aphasias and Aprosodias Aprosodias"' Aphasias Comprehension Spontaneous of Emotional Prosodie Prosody and Prosodie Gesturing Gesturing Repetition Comprehension Fluency Repetition Comprehension Comprehension Reading Motor Sensory Global Conduction Transcortical motor Transcortical sensory Mixed transcortical Anomic (alexia with agraphia) Good Poor Good Poor Poor Poor Poor Poor Good Good Good Good Good Good Poor Good Poor Good Poor Poor Good Good Poor Poor Good Poor Poor Poor Poor Good Poor Good Good Good Poor Poor Poor Poor Poor Good Poor Good Good Good Poor Good Good Poor Good Poor Good Poor Poor Good Good Poor Poor Good Good Poor Poor Poor Good Poor been hypothesized; the others have been described." Motor, sensory, global, and with lesions in the left hemisphere known to cause homologous aphasias.26 tures in patients with right-sided as opposed to left-sided cerebrovascular accidents, they were not certain if their findings were a consequence of lesion location or a specific psychiatric problem. With the recent elucidation of more precise anatomical knowledge about the emotional func¬ tions of the right hemisphere,'41s disentangling the rela¬ tionship of depression and brain damage now may be feasible. The following case reports illustrate the vexing problems of recognizing depression in brain-damaged patients, and clinical guidelines for diagnosing the depressive syndrome in these patients are offered. In those cases where a well-defined anatomical lesion could be documented by computed tomography (CT), inferences about the neuroan¬ atomical basis of the depressive syndrome are made. PATIENTS AND METHODS The patients to be described primarily were referred to us because of behavioral problems. None had been given a diagnosis of depression prior to referral, suggesting that depression may be largely unrecognized in the brain-damaged population. All patients underwent a dexamethasone suppression test (DST), since this test has been shown to be a valuable neuroendocrine marker for endogenous depression.'2 The DST was performed by giving dexamethasone (1 mg orally) at midnight and measuring serum cortisol levels at 8 am, 4 pm, and 11 pm the following day. Nonsuppression was defined by one or more cortisol levels in excess of 5 ug/mL after dexamethasone. Sixty percent to 80% of psychiatric inpatients with depression are reported to have nonsuppression that reverts to normal suppres¬ sion with effective antidepressant treatment.36 All patients underwent CT scanning, which constitutes the anatomical data base for our inferences about the neurological basis of depression (Fig 1 through 4, Table 2). To date, our clinicopathological correlations suggest that the signs and symptoms of depression may be divided into four major categories. (1) The verbal-cognitive set consists of the actual words and syntactical arrangements used by the patient to convey to the examiner, through propositional language, the depth and severity of his dysphoria, his "understanding" of the causes for his mood, and his notions about himself and his future. (2) The affect category refers to the facial expressions, body postures, and emotional components of speech from which the clinician infers the depth, nature, and severity of the patient's dysphoria (mood). (3) The category of vegetative behavior includes the changes in appetite, sexual drive, sleep, and motivational level that accompa¬ ny most depressions. These also help the clinician infer the depth and severity of a patient's dysphoria. (4) Mood refers to the actual inner feelings of dysphoria that a patient experiences, which the clinician can only infer from the patient's verbal-cognitive set, affect, and vegetative behavior. Although there is a general consensus that "unipolar depres¬ sion" comprises a heterogeneous group, controversy remains as to which subclassification system is most valid.IT Klein'* and others have suggested that unipolar depression can be categorized into at least the following two distinct groups: endogenous (endogenomorphic) depression, with derangements in the pleasure-reward system, and nonendogenous depression, with an underlying pessi¬ mistic personality disorder. Recent studies using biological mark¬ ers, such as the DST and the sleep EEG, have supported this endogenous-nonendogenous dichotomy in unipolar patients.™ Endogenous depressions are characterized by a distinctly unreactive mood with pervasive anhedonia, marked vegetative signs, abnormal biological markers, and a good response to antidepres¬ sants. Nonendogenous depressions are characterized by fewer and less severe vegetative signs, fewer if any abnormalities in biolog¬ ical markers, a variable response to antidepressants, and, possibly, '"~42 a better response to selected psychosocial interventions. Typi¬ cally, patients with endogenous depression provide verbal-cogni¬ tive constructs that are insufficient to account for the depth and severity of their dysphoria, whereas the verbal-cognitive con¬ structs provided by patients with nonendogenous depressions appear better matched to their dysphoria in both timing and '" intensity.3*·" REPORT OF CASES '" Case 1.—Left Hemisphere's Contribution to Verbal-Cognitive Set.— An 81-year-old woman was referred for neurological evaluation six weeks after having a cerebrovascular accident that involved the left superiortemporal and inferior parietal lobes with a resulting Wernicke's aphasia. Following the cerebrovascular accident and a brief hospitalization, she began to have severe behavioral changes that subsequently led to her dismissal from two nursing homes. Extreme irritability, negativism, and explosive outbursts were noted. She was unwilling to assist in routine personal hygiene, refused food, displayed no interest in social interaction, and was observed to continuously stare out the window. She either ignored or displayed marked annoyance to the presence of music, people, and television and often refused to speak to anyone. Because of her intolerable behavior, she was given haloperidol. According to her daughter, however, the medicine only "quieted her down" but still left her "seething underneath." This clinical course is very atypical for patients with chronic Wernicke's aphasia. Such patients are usually excessively verbal, cooperative, and easygoing. They maintain personal hygiene, seek social interaction, and eat with normal gusto. Extensive interviews with the patient's daughter revealed a history that previously had been overlooked that strongly sug¬ that the patient had an endogenous depression. Until the death of her husband four years prior to the stroke, the patient had been energetic and outgoing, with a wide range of interests. She then moved from New York to Dallas and was noted to have a marked and progressive reduction in social interactions with a loss gested Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 Fig 1.—Computed tomographic scan of patient 1. Infarction (arrowheads) involves left posterosuperior temporal lobe and posteroinferior parietal lobe (Wernicke's area). Fig 2.—Computed tomographic scan of patient 2. Infarction (arrowheads) involves right striatum and subcortical white matter of frontal and anterior parietal opercula. Although cortical involvement is not yet evident (six days after stroke), deep distri¬ bution of infarction implies occlusion of superior division of right middle cerebral artery. of general interest. She became notorious for refusing to partici¬ pate in family activities and tended to talk negatively about herself, her life, and her future. Anorexia, with a 13.5-kg (30lb) weight loss, was evident prior to the stroke. Midnocturnal and early-morning insomnias, memory problems, and poor concentra¬ tion also were reported. There was no previous history of depres¬ sion, mania, or other psychopathological syndromes. On neurological examination, the patient had fluent but severely paraphasic speech with very poor comprehension and repetition. She was inattentive and often stared off into space while being interviewed. She was totally unable to report her internal affective state, the presence of vegetative symptoms, or her preferences for food, television, and other environmental activities. Mild psychomotor retardation was evident, and she was generally irritable and uncooperative. A repeated CT scan (Fig 1) revealed only one lesion, as previously described, and a dementia workup produced negative results. Since the history provided by the daughter strongly suggested that a depressive syndrome had been present prior to the stroke, a diagnosis of endogenous depression was made (Table 2). Sleep-onset, midnocturnal, and early-morning insomnias and marked anorexia were confirmed by hourly nursing observations, and a DST revealed nonsuppression with a 4 pm serum cortisol level of 11.6 fig/dL after dexamethasone. Gradually increasing dosages of amytriptyline hydrochloride at bedtime (up to 50 mg/day) were administered over two weeks, and her sleep pattern, food ingestion, general mood, social interaction, and interest in personal hygiene were monitored carefully. Within three weeks, her sleep pattern normalized and she began to eat with enthusiasm, gaining 2.25 kg (5 lb). She became far less irritable and negative and showed significant improvements in general interest, social interaction, and attention. The neurologi¬ cal examination now indicated a classic Wernicke's aphasia. The patient became talkative, smiled frequently, and seemed inter- Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 Fig 3.—Computed tomographic scan of patient 3. Two infarctions are seen (arrow¬ heads); one involves right midsuperior parietal lobe and second involves right inferior frontal lobe. Sulcal widening and ventricular enlargement are consistent with patient's age. Fig 4.—Computed tomographic scan of patient 4. Infarction (arrowheads) involves most of right frontal and parietal lobes, with extension into posterosuperior tem¬ poral lobe. There is also infarction of left anteromedial frontal lobe. Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 ested in attempting to converse and communicate even though her aphasia made coherent discourse impossible. After two weeks with amitriptyline treatment, a repeated DST showed normal suppression. Because her behavior had improved so markedly, the patient's family, who previously had insisted that she go to a nursing home, decided to care for her in their own home! At follow-up four months later, a continued excellent therapeutic response to amitriptyline was noted. Clinicopathological Correlations.—This case illustrates the diffi¬ culty of diagnosing depression in a patient with severe compre¬ hension deficits from aphasia. Since taking a history was impossi¬ ble, we had to depend on direct observations of the patient's behavior and secondary sources of historical information. Although haloperidol partially controlled her behavior, a more appropriate and specific treatment led to substantially greater therapeutic effects. The DST corroborated our psychiatric diagno¬ sis. As a consequence of the Wernicke's aphasia, patient 1 lost her verbal-cognitive capacities and was unable to communicate her depression to the examiner and family through propositional language. Nevertheless, she continued to display affective behav¬ iors (negativism, irritability, social withdrawal, and psychomotor retardation) and vegetative behaviors (insomnia, anorexia, and weight loss) consistent with endogenous depression. These severe behavioral abnormalities are virtually never seen in the recovery phase of uncomplicated Wernicke's aphasia. Therefore, we can infer that the patient's internal feeling state (mood) was one of dysphoria even though there could be no validation of this inference through propositional language. In contrast to patients with Wernicke's aphasia, patients who have Broca's aphasia from lesions involving the left inferior frontoparietal region are characterized by nonfluent speech and poor repetition with relatively preserved auditory comprehen¬ sion.2" Clinical experience suggests that when persons with Broca's aphasia are depressed, they usually will have a mood-congruent verbal-cognitive set even though they have great difficulty expressing it through propositional language. The examiner can elicit this verbal-cognitive set by either questioning the patient using a yes-no or multiple-choice interview format" or by using standard rating scales,'1 such as the Hamilton Rating Scale for Depression,'" so long as reading and/or auditory comprehension are relatively intact. These observations coupled with those from patient 1 suggest that the verbal expression of a depressive verbal-cognitive set requires both Broca's and Wernicke's areas, whereas the internal formulation of a depressive verbal-cognitive set depends on Wernicke's area. Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 Case 2.-Right Hemisphere's Contribution to Depressive Affect.—A 34-year-old man was admitted to the hospital following the rapid onset of left hemiplegia. He had no previous history of neurologi¬ cal, psychiatric, or medical problems other than hypertension, for which he intermittently took methyldopa. Initial examination revealed a dense left hemiplegia with hemisensory loss and a left hemianopsia with head and eyes deviated to the right. He was lethargic and dysarthric but not aphasie. The presumptive diagno¬ sis was a cerebrovascular accident in the distribution of the right middle cerebral artery. A CT scan revealed a right inferior frontoparietal stroke (Fig 2, Table 2). The lethargy, hemianopsia, dysarthria, and head and eye devia¬ tion cleared over seven days. A flat monotonie voice, devoid of emotion, and the absence of spontaneous gesturing were noted. He comprehended the emotional components of language and gesture but could not repeat sentences with affective variation. These deficits in emotional behavior are consistent with the newly described syndrome of motor aprosodia that is a consequence of right frontal opercular brain damage." ' About 17 days later, he began to show intermittent and uncon¬ trollable crying that was an "all-or-none" phenomenon. The patient complained that he had no control over these emotional outbursts and that they did not necessarily correlate with his mood or the content of his conversation. (These findings in brain¬ damaged patients are known as pathological laughing and cry¬ ing.11) Nevertheless, in his totally flat and aprosodic voice he stated that he felt "hopeless," "depressed," and "helpless," but he denied having suicidal thoughts. Over the next two weeks, howev¬ er, he reported suicidal ideation and feelings of being severely depressed, with mild midnocturnal insomnia and a decrease in general interest. No appetite or weight disturbances were not¬ ed. On mental status evaluation, this cooperative and fully oriented man was given to intermittent crying for which he often had no explanation. He ascribed his sad mood to his physical impairment and the occupational consequences of his stroke. He was hopeless and blamed himself for the stroke, since he had complied poorly with his antihypertensive treatment. Attention and concentration were intact. The patient had no dysarthria, aphasia, or generalized intellectual impairment. However, his speech was aprosodic and unaccompanied by gesturing. This aprosodic-agestural state, a consequence of the neurological lesion, gave the patient a flat¬ tened affect, making it impossible to assess accurately the depth and strength of his internal mood through nonverbal behavior. He did not have a depressive affect. Because of his suicidal ideation, imipramine hydrochloride (150 mg/day) was begun. A complete remission of all depressive symptoms occurred within 14 days, although his flattened affect and motor aprosodia remained unchanged. A DST prior to treat¬ ment revealed normal suppression. Clinicopathological Correlations.-This patient probably had a "reactive" (nonendogenous) depression as he became aware of the consequences of his neurological disability. Although his symp¬ toms remitted in association with imipramine treatment, he also confronted and resolved some of the issues and concerns gener¬ ated by the stroke. Specific neurological lesions in the right hemisphere appear to result in aprosodic-agestural speech with retained comprehension of the affective components of language,'" as was encountered in this patient. Because these patients have a flattened affect, the physician is unable to utilize affective behaviors as clues by which to infer the patient's internal mood state. Consequently, clinicians may disbelieve the patient's propositional statements about their mood, since such statements will be delivered in an unconvincing, flat, monotone devoid of affective qualities. To avoid overlooking depression in these patients, strict attention must be paid to their propositional language statements about inner mood. The agestural-aprosodic state that follows right frontal brain lesions, however, is a distinct neurological entity that should not be confused with or misinterpreted as a depressive affect." Although patient 2's verbal-cognitive constructions were related in time and intensity to his reports of dysphoria, he did not show a typical depressive affect, even though his depression was severe enough for him to contemplate suicide. Thus, this case suggests that the right inferior frontoparietal lobes are crucial for organiz" ing and modulating the affective behaviors that typically accom¬ pany depression. Apparently, however, they are not critical for the development of either the verbal-cognitive set or the dysphoric mood. Case 3.—Right Hemisphere's Contribution to Depressive Affect.—A 70-year-old man was hospitalized following a syncopal episode. Two years previously, he had sustained an embolie stroke second¬ ary to atrial fibrillation that caused a left hemiparesis and hemisensory loss that resolved over time. On admission, he had a grade 2/6 apical pansystolic murmur, atrial fibrillation, and mild weakness of the left upper extremity. One week later, he had a second right hemispheric stroke that caused a severe left hemiple¬ gia with sensory loss. He was noted to have acute delirium with hallucinations. After Va weeks, however, he was only mildly inattentive and could do six numbers forward on testing imme¬ diate recall. Recent and remote memory were intact. There was no aphasia, but his speech was aprosodic and agestural, which pro¬ duced a flattened affect. Prosodie repetition was very poor, but comprehension of the emotional components of speech and gesture was good and he was given the diagnosis of motor aprosodia.11 The patient had little insight into his deficits and denied depression. During examination, however, he had multiple and uncontrollable laughing and crying spells that had an all-or-none quality. He denied feeling sad or happy during these episodes, which is consistent with the syndrome of pathological laughing and crying. A CT scan confirmed an "old" right superior parietal stroke and a new right inferior frontal stroke (Fig 3, Table 2). Over the next few weeks, marked insomnia, severe anorexia, lack of interest in activities, and the gradual admission of dysphoria ("I feel blue") were noted. Although he denied suicidal thoughts, he showed guilt and self-blame. He was preoccupied with the whereabouts of his wife, whom he believed was unfaithful and had run off with another man. He also spoke of a severe injury to the right eye that he incurred during World War II and began to recount harrowing war experiences. An independent history then was obtained from his wife, who had been visiting him daily in the hospital. None of his WW II ruminations or his alleged eye injury was substantiated. Five years prior to his stroke, the patient had retired from his usual occupation because of difficulties with concentration and memory. He lost virtually all interest in previous hobbies and activities and began watching television all day. He cried frequently and talked intermittently about feeling worthless. Following WW II, he apparently had had a severe episode of depression that would have met Research Diagnostic Criteria28 for a major depressive disor¬ der. His wife recalled that during that episode he also was preoccupied with an alleged severe injury to the right eye and harrowing war experiences. This history coupled with the current symptoms of dysphoria, anorexia, and insomnia led us to the diagnosis of endogenous depression, psychotic type. A DST revealed nonsuppression, with a serum cortisol level of 13.6 pg/dh at 4 pm after dexamethasone. Subsequent clinical deterioration from two episodes of ventricular tachycardia with cardiac arrest and a third stroke precluded the initiation of antidepressant treatment. Clinicopathological Correlations.-This case illustrates the critical role of an independent history in substantiating and clarifying clinical observations. The difficulty in correctly gauging the depth of a patient's dysphoria in the presence of a motor aprosodia is again apparent. Nevertheless, patient 3 was able to describe his mood through propositional language and displayed vegetative signs of depression. Although initially his inattention and delu¬ sional state were attributed to his brain lesions, it became obvious, with further history, that he had an endogenous depression, psychotic type. The abnormal DST further substantiated our clinical impression. As with patient 2, when this patient spoke about his dysphoria and delusions (verbal-cognitive set), he did not evidence moodcongruent affective behaviors that are typically associated with endogenous depression. The brain lesion associated with this patient's loss of affective behavior is best correlated with his second stroke involving the right inferior frontal region. Thus, this area in the right hemisphere appears to be crucial for modulating the depressive affect, but it is not necessary for the formulation or expression of the depressive verbal-cognitive set, the dysphoric Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 mood, or the vegetative signs associated with endogenous depres¬ sion. The previous negative inferences also hold true for the right midsuperior parietal area, which was destroyed by the first stroke. Case 4.—Right Hemisphere's Contribution to Depressive Affect and Verbal-Cognitive Set.—A 35-year-old woman was in good health until the acute onset of a subarachnoid hemorrhage. A left internal carotid berry aneurysm was clipped successfully, but she sustained a postoperative left hemiplegia and hemisensory loss due to an ischemie infarction in the right hemisphere. Six months later, she was admitted to the hospital for rehabilitation. She was noted by the rehabilitation personnel to be very difficult to work with since she was negative, irritable, and either unwilling or unable to cooperate in her physical therapy. On examination, she was alert, awake, and attentive. Recent and remote memory and proverb interpretations were intact. She was aware of her neurological deficits but had no true emotional reaction to them. She was not aphasie but had aprosodic-agestural speech that gave her a flattened affect. Prosodie repetition was relatively preserved, but comprehension of the emotional compo¬ nents of speech and gesturing was poor. (This constellation of findings is consistent with the syndrome of mixed transcortical aprosodia, which, in this patient, probably evolved from a global aprosodia.17) She had a severe left hemiplegia, global left hemisen¬ sory loss, and urinary incontinence. A CT scan showed a large right supra-Sylvian infarction with extension into the superiorposterotemporal lobe and a small left anterofrontal infarction (Fig 4, Table 2). She specifically and consistently denied every sign and symptom of depression, including dysphoria, anorexia, anhedonia, easy fatigability, impaired thinking, suicidal ideation, and guilt both during hospitalization and in clinic follow-up. Three months after her initial evaluation, the patient's mother accompanied her to clinic, and a radically different history was obtained. Her mother had noted over the previous six months that the patient had marked anorexia ("eats like a bird"), severe insomnia, and excessive irritability with verbal and physical assaults of her children, behaviors quite unlike her former self. She often required several people to assist her in and out of bed, a degree of dependency that was far out of proportion to her neurological deficits. She avoided social interactions and preferred to remain in her room by herself. In addition, a consistent history of pathological affect regulation characterized by uncontrollable laughing or crying at the slightest stimulus was obtained. Review of her hospital record substantiated a 27-kg (60-lb) weight loss, and a diagnosis of endogenous depression was made. A DST revealed nonsuppression, with a serum cortisol level of 12.6 µg/dL at 4 pm after dexamethasone. Treatment with desipramine hydrochloride (125 mg at bedtime) led to marked resolution of her irritability, dependency, anorexia, and insomnia. She became socially interactive and motivated to assist in her rehabilitation. Her energy level was increased, and she gained 9 kg (20 lb). Her pathological laughing and crying completely abated. A repeated DST revealed normal suppres¬ sion. Clinicopathological Correlations.—The absolute necessity of obtaining a history from nonpatient sources in neurologically impaired patients is underscored once again. This is especially important when the patient evidences "denial of illness,"48 a well-documented consequence of right-sided brain lesions. This patient denied every sign and symptom of depression, even when confronted with her mother's observations concerning her sleep, appetite, weight change, and irritability. To further complicate matters, the patient had a mixed transcortical aprosodia that caused her to have difficulty in both expressing and comprehend¬ ing the affective components of language. The value of the DST as an adjunct to diagnosis also is demonstrated. In addition to denying all the signs and symptoms of depression, the patient spoke indifferently about her neurological deficits, a finding that persisted even after successful treatment of her depression. The only occasion when she spoke with notable concern was during a discussion about her brother, who two years prior to her stroke had been arrested and sent to prison for armed robbery. The patient had been very close to him and was devastated by his trial and incarceration. She spoke of being "sad," "broken¬ hearted," "torn apart," and "disappointed," even though her voice patterns and behavior were without affect. Yet, whenever she spoke about her devastating neurological deficits she never dis¬ played a similar affectively charged verbal-cognitive set. This patient provides us with the greatest insight into the anatomical basis of depression. Since she denied every sign and symptom of depression, including vegetative features, her verbalcognitive set ("I'm not depressed... I feel good") actually was totally irrelevant to her inferred internal mood state and depres¬ sive disorder. In contrast, when she talked about her brother's incarceration, her expressed verbal-cognitive construction of this event was affectively appropriate, demonstrating that her verbalcognitive capacities were intact. This striking dissociation can be explained by the fact that when she initially experienced the pain and loss of her brother, her brain was intact. Therefore, any affective information processed in the right hemisphere could be transferred across the corpus callosum to the propositional lan¬ guage areas in the left hemisphere, which could then construct and ultimately learn an appropriate affectively charged verbal-cogni¬ tive set. On the other hand, when we questioned her about her internal feelings concerning her neurological deficits or her mood state, we actually were interviewing an isolated left hemisphere that could not access the emotional information stored and processed in her damaged right hemisphere. Thus, her left hemi¬ sphere gave us irrelevant, misleading, and affectless propositional responses. This is similar to some of the experiences that Gazzaniga and Le Doux4" report when dealing with patients with "split brain." In cases 2 and 3, the right hemisphere lesion spared the posterosuperior temporal-posteroinferior parietal lobes (the homologue of Wernicke's area in the right hemisphere),'7 and both patients displayed a mood-congruent verbal-cognitive set when depressed. However, in case 4 the neurological damage involved the right posterosuperior temporal-posteroinferior parietal lobes, suggesting that this area is crucial for modulating and communi¬ cating the affective set via the corpus callosum to Wernicke's area in the left hemisphere for the formulation of a mood-congruent verbal-cognitive set (case 1), and to the right inferior frontal region for expression of the depressive affect (cases 2 and 3). We cannot be sure that this patient was truly dysphoric, since her left hemisphere denied being dysphoric, but was her right hemisphere dysphoric? Without ways to verbally question her right hemisphere, we only can infer dysphoria based on her striking clinical and biological evidence for an endogenous depres¬ sion. The same question and inferential process also applies to patient 1, whose Wernicke's apahsia, which was secondary to a left hemisphere lesion, also prevented us from directly ascertaining if she was dysphoric. A final clinicopathological correlation in this patient concerns the anatomical origin of endogenous depression. If her depression, with its inferred dysphoria, originated in her relatively intact left hemisphere, then one would have expected the patient to have provided a mood-congruent verbal-cognitive set. Since this was not the case, we suspect that her depression was initiated by structures in the right hemisphere that were not involved by the stroke. Exactly which right hemisphere structures, however, can¬ not be ascertained in this patient using clinicopathological corre¬ lations. A likely candidate, based on ablation and stimulation studies in humans and animals, might be the amygdala,50,2 with its powerful limbic and hypothalamic connections and known involvement in the modulation of inner emotional states. Case 5.—Closed Head Trauma and Depression.—A 38-year-old referred for neuropsychiatrie evaluation two years after a motor vehicle accident that had rendered her unconscious and semicomatose for approximately 2Vfe weeks, with gradual recovery. The EEG evidence of slowing over the right cerebral hemisphere suggested a destructive process, but a CT scan only demonstrated slightly enlarged ventricles. She was discharged approximately two months later with residual neurological deficits of motor incoordination in her extremities, urinary incontinence, inattention, and a euphoric affect. One year after her accident, she began to show a progressive decline in her behavior and intellectual abilities. Temper tan¬ trums, marked irritability, extreme uncooperativeness, and refus¬ al to participate in household activities were noted. Thioridazine hydrochloride (10 mg three times a day) was begun, but it only woman was Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 blunted some of her more extreme outbursts. We first examined this patient two years after the accident. Further history disclosed that she had pathological laughing and crying spells for the past year. Both the family and the patient concurred that she had no trouble sleeping or eating and had, in fact, gained 13.5 kg (30 lb). The family's report of irritable behavior, however, was denied by the patient. She did admit that for approximately one year she had experienced continuous sad¬ ness and helplessness without suicidal ideation. On admission, she was awake, alert, and cooperative, with intact attention, immediate recall, and recent and remote memories. Proverb interpretation was very concrete. Her affect was extremely labile, swinging from uncontrollable laughing to crying that was unrelated to the context of the discussion. In between the spells of pathological affect regulation, she spoke in a flat monotone without.prosodie inflection or gesturing, which created a flattened affect. Insight into her illness was good even though she had little insight into the impact of her behavior on her family. She was not aphasie, but her speech was aprosodic, with poor ability to repeat with emotional variation. Comprehension of the verbal and gestural components of emotional language was rela¬ tively good, which is consistent with the syndrome of motor aprosodiaf1 Copying figures, handwriting, calculations, and rightleft orientation were intact. The cranial nerves were normal. Motor examination revealed mild flexion weakness at her left hip, with some spasticity in her left lower extremity that was greater than in her right lower extremity. The reflexes were slightly increased on the left side, with a left upgoing toe. Her sensory examination was intact, but her gait was wide based and slightly spastic. A repeated CT scan showed mild enlargement of the ventricles, with definite atrophy of the prefrontal regions bilaterally. A DST showed normal suppression. A regimen of nortriptyline hydrochloride treatment (50 mg at bedtime) was begun; after 2lk weeks, the dose was increased gradually to 100 mg at bedtime. She was seen in follow-up one month later and had sustained a remarkable improvement in her behavior, with total resolution of her dysphoria. She no longer demonstrated pathological laughing or crying, irritability, or insomnia. She was once again helping around the house and taking on responsibilities. Her neurological deficits remained un¬ changed. Clinicopathological Correlations.—This case again demonstrates the value of interviewing family members to help clarify the clinical history and was included to illustrate that the strategies for diagnosing depression in patients with closed head trauma are the same as for patients with strokes (cases 1 through 4). Detailed clinicopathological correlations, however, could not be made because the CT scan did not define the extent of her brain lesions, which is often the case in patients with closed head trauma. COMMENT Diagnosing Depression in Brain-Damaged Patients While a transient sense of discouragement, hopelessness, sadness, or demoralization may follow any neurological injury, some patients actually may have a true biological depression." '"" The depression may develop following the neurologic insult (cases 2, 4, and 5) or preexist the insult (cases 1 and 3). The diagnosis and appropriate pharmaco¬ logical treatment of the depression, however, is critical since untreated depression may impair recovery from the neurological damage (cases 1, 2, 4, and 5) or may interfere with rehabilitation (cases 2 and 4).M Furthermore, neuro¬ leptics may be administered inappropriately in attempts to reduce behavioral problems that arise from the combina¬ tion of brain damage and depression (cases 1 and 5), thus leaving the depression untreated. Several bedside clues to the diagnosis of depression in neurologically impaired patients are suggested by these cases (Table 3). Erratic, abnormal, or poor recovery from the neurological insult (cases 1, 3, and 4), a failure to participate appropriately in a rehabilitation program Table 3.—Clues for Diagnosing Depression in Neurologically Impaired Patients First signs of depression in brain-damaged patients may in¬ clude (a) poor or erratic recovery from neurological insult, (b) failure to cooperate in rehabilitation, (c) "management" difficulties, or (d) clinical deterioration from previously sta¬ ble neurological deficit Interview both patient and family for signs and symptoms of depression, paying special attention to vegetative signs; some brain-damaged patients may deny many or all of signs and symptoms of depression and still be depressed Presence of pathological laughing and crying In brain-dam¬ aged patients who do not have pseudobulbar palsy sug¬ gests that they have depression Aprosodic voice patterns, lack of emotional gesturing, and eu¬ phoric behavior may occur following R hemisphere lesions; these behavioral patterns are independent of depression; therefore, do not disregard patient's propositional language statements about dysphoria, guilt, hopelessness, and sui¬ cide even if affective behaviors are ¡ncongruent with state¬ ments; on the other hand, flattened affect should not be misinterpreted as depressive affect Abnormal dexamethasone suppression test is valuable biologi¬ cal marker to help establish existence of endogenous de¬ pression in some brain-damaged patients Trial of antidepressant medication is indicated in all suspected cases; neuroleptics may subdue behavioral problems in brain-damaged patients with depression, leaving depression untreated (cases 2 and 4), or worsening of a neurological deficit after recovery (case 5) may be the first signs. The clinician then must obtain a careful psychiatric history to determine if the patient is depressed. This history should be obtained from both the patient and other family members since neurological insults may prevent accurate history taking, especially when the patient has an aphasia with compre¬ hension deficits (case 1) or denial of illness (case 4). In some cases, vegetative signs may be the only clue to a depres¬ sion, and these signs should be elicited carefully from the family and, if need be, looked for by nursing personnel (cases 1 and 3). Aprosodic voice patterns with lack of gesturing fre¬ quently are encountered in patients with right hemisphere lesions,'"'7 causing them to have a flattened affect. In these instances, the patient will be unable to modulate his voice and gestures appropriately. Such patients may report dysphoria in an unconvincing, monotonous, and aprosodic voice (cases 2, 3, and 5), which readily could mislead the clinician into minimizing or discounting verbal reports of dysphoria. Conversely, aprosodic voice patterns and lack of gesturing should not be misinterpreted as indicating the presence of depression. If the patient's right hemisphere lesion is large, causing deficits in comprehension of affec¬ tive language and denial of illness, then verbal reports concerning mood, depression, affective behavior, or vege¬ tative symptoms may be confabulatory and totally inaccu¬ rate (case 4). A striking finding in this series of patients is that pathological laughing and crying (cases 2, 3, 4, and 5) may constitute a valuable behavioral marker for depression in those brain-damaged patients who have a flattened affect secondary to lesions involving the right inferior frontal lobe. Pathological affect regulation is usually the result of bilateral brain lesions causing pseudobulbar palsy with pseudobulbar affect,2" but none of our patients had pseudo¬ bulbar palsy. Although our patient's pathological laughing and crying totally abated with antidepressant treatment (cases 2, 4, and 5), suggesting that depression was the critical factor, both conditions (right-sided brain damage and depression) are necessary since neither alone appears Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 Fig 5.—Neuroanatomical model of endogenous depression. Numbers and model are explained in Table 4. Table 4.—Neuroanatomical Model for Signs and Symptoms of Endogenous Depression Area* 1 (Broca's area) (Wernicke's area) Neurological Syndrome Postulated Function in and Function Endogenous Depression Expression of depressive verbal-cogni¬ tive set through propositional lan¬ Motor (Broca's) aphasia; organizes mo¬ tor output for propositional language guage Internal formulation and ultimate ex¬ pression of depressive verbal-cogni¬ tive set through propositional lan¬ guage Expression of depressive affect through affective components of language and behavior Internal formulation of "depressive-af¬ fective set" that allows, via corpus callosum, Wernicke's area to formu¬ late mood-congruent depressive ver¬ bal-cognitive set and area 3 to ulti¬ mately express mood-congruent de¬ pressive affect Unknown structure or structures essen¬ tial for initiating and/or modulating critical features of endogenous de¬ Sensory (Wernicke's) aphasia; involved in comprehension of propositional components of language Motor aprosodia; organizes motor out¬ put for affective components of lan¬ guage and behavior Sensory aprosodia; involved In compre¬ hension of affective components of language and behavior Not known pression (eg, dysphoria, vegetative behavior, hypothalamic dysfunction) 'Areas are shown in Fig 5. sufficient to precipitate pathological affect regulation. These cases also illustrate the value of the DST as an adjunct in the diagnosis of depression in brain-damaged patients. Three of our five patients demonstrated DST nonsuppression, which was consistent with their diagnosis of endogenous depression (cases 1, 3, and 4). Furthermore, normalization of the DST in cases 1 and 4 occurred in association with clinical remission, similar to the reported experience in patients without brain damage with DST nonsuppressing depressions.™ In brain-damaged patients with normal DSTs but in whom the diagnosis of depression is suspected, antidepressant medication still may be of benefit (cases 2 and 5). The complicated and often confusing clinical picture of depression in neurologically impaired patients easily can be ascribed to nonspecific agitation, situational adjust¬ ment reaction, or the neurological lesion. To avoid these errors and to institute proper treatment, psychiatrists, neurologists, and rehabilitation specialists must become aware of the protean manifestations of depression in brain-damaged patients and apply the correct techniques for making an appropriate diagnosis. Exclusive reliance on standard psychiatrie evaluations that depend heavily on criterion signs and symptoms found in depressed patients without brain damage may lead to diagnostic errors. Neuroanatomical Basis of Depression Figure 5 and Table 4 provide a model for the functionalanatomical organization of the signs and symptoms encountered in endogenous depression that we have de¬ rived from the case reports (Table 2). As with all models, it should not be construed as either proved or final but rather as a working hypothesis with heuristic and predictive power that will be modified as further clinicopathological data accrue. For instance, there is a group of propositional language disorders, known as the transcortical aphasias2" (Table 1), in which the causative lesions usually lie outside the classic peri-Sylvian language areas (eg, Wernicke's and Broca's areas). Patients with transcortical aphasia all have in common the preserved ability to repeat sentences that are spoken to them, which is in distinct contrast to the aphasias associated with peri-Sylvian lesions (Table 1). Although Wernicke's area usually is considered to be responsible for the comprehension of propositional Ian- Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 guage, the existence of transcortical sensory aphasia strongly suggests that Wernicke's area is only a crucial way station for comprehension and that comprehension is actually a function of the cortex surrounding Wernicke's area.2""'"' Similarly, even though our initial study has implicated Wernicke's area as being a crucial structure for the formation of a depressive verbal-cognitive set, it is possible that the areas truly responsible for formulating the depres¬ sive verbal-cognitive set will be shown to reside outside Wernicke's area. Nevertheless, we believe our proposed model will prove valuable in unraveling the neuroanatomi¬ cal basis of depression since it is internally consistent with the known functional-anatomical organization of the prop¬ ositional·'" and affective17 components of language in the brain. The model also receives indirect support from Dimond's""'"'7 observations concerning vigilance and con¬ sciousness in patients with partial and complete surgical commissurotomies. He hypothesizes that the brain circuits responsible for conciousness in each hemisphere reside "towards the back of the brain" and are unified via the splenium of the corpus callosum. Similarly, based on cases 1 and 4, we hypothesize that Wernicke's area and its homo¬ logue in the right hemisphere are responsible for formulat¬ ing the verbal-cognitive and affective sets in depression, which then interact with each other via the corpus callosum (Fig 5). Recent anatomical data in humans suggests that these particular brain areas are connected through the splenium!"' The model is also consistent with Flor-Henry's1 '··"·" hy¬ pothesis that mania, depression, and schizophrenia relate to functional alterations in the frontotemporal regions of the brain. The interaction of these systems, via the corpus callosum, determines the type of psychiatric disorder encountered. For instance, if a depression (right hemi¬ sphere) triggers changes in the left frontotemporal limbic system, then a psychotic melancholia with paranoid delu¬ sions may occur. Three points about our findings and model deserve emphasis. First, depression, even though it is psychiatrically classified as a disorder of affect, is neither a "right"- nor a "left"-hemisphere syndrome; both hemispheres partici¬ pate in the depressive syndrome, each modulating certain signs and symptoms. This concept is important to under¬ stand since we are now on the verge of applying powerful new technologies, such as positron emission tomographic (PET) scanning'1" and Xenon Xe 133 blood flow scanning,'" to help elucidate the anatomical-metabolic derangements that accompany depression. Proper interpretation of these types of data will depend on understanding the functional- anatomical organization of the depressive syndrome. To this point, it is interesting that the preliminary studies using the PET"" and '"Xe scanning techniques'" have demonstrated not only asymmetrical but also bilateral metabolic changes in the brains of depressed patients. Perhaps some of the controversies and interpretations surrounding recent electrophysiological and EEG findings showing right-vs-left or left-vs-right asymmetries in depressed patients4"2" are resolvable if we accept that both hemispheres participate in the depressive syndrome. Although our initial findings suggest that endogenous depression may be initiated by structures in the right hemisphere, one should not make the assumption that this will hold true for nonendogenous depressions. Since nonen¬ dogenous depressions seem to be characterized by verbalcognitive constructions (left hemisphere) that are better matched in timing and intensity to the patient's dyspho¬ ria,'8" perhaps, nonendogenous depressions originate in a left hemisphere structure or structures, which then secon¬ darily effect the areas in the right hemisphere that modulate affective behavior. Such an anatomical basis might account for the better matching of verbal-cognitive constructs to dysphoria in nonendogenous as compared with endogenous depressions. The reason for the poorer matching of verbal-cognitive constructs to dysphoria in endogenous depression can be explained if endogenous depression does, in fact, arise from right hemisphere structures that secondarily affect areas in the left hemi¬ sphere that are critical in formulating the verbal-cognitive set. Until further clinicopathological correlations are made, however, the exact anatomical origins of both endogenous and nonendogenous depressions can only be hypothe¬ sized. Last, if the model is correct, then it should predict the existence of certain clinical syndromes not described pre¬ viously. Although a number of syndromes can be predicted, a crucial patient for validating our basic hypotheses would be one who has both an endogenous depression and a sensory aprosodia secondary to a lesion involving the homologue of Wernicke's area in the right hemisphere (area 4 in Fig 5). Neurologically, the patient would be unable to either repeat or comprehend the affective com¬ ponents of language. The affective components of lan¬ guage and behavior during spontaneous communication, however, would be retained, although mismatching of the affective to propositional components of language might occur. In fact, this patient may even appear euphoric and unconcerned, as has been reported previously in a patient with sensory aprosodia.17 Because of the pivotal anatomical area destroyed (Fig 5), the usual unfolding of the signs and symptoms comprising endogenous depression should be altered greatly. The patient should not display a depressive verbal-cognitive set (even though his verbal-cognitive capacities are preserved) or affective behaviors consistent with either depression or a flattened affect. The only clinical manifestations of the endogenous depression should be marked changes in vegetative behavior. Treat¬ ment with antidepressants, as in case 4, should resolve his vegetative behavioral abnormalities without changing his "inappropriate" verbal-cognitive set or affective behav¬ ior. In summary, this article presents an initial inquiry into the neurology of depression utilizing the technique of clinicopathological correlations. A more thorough under¬ standing of the biology of depressive disorders is likely to ensue by combining our data and hypotheses with future clinical, pharmacological, physiological, and neurochemical investigations. This in turn should establish firmly that depression is primarily a disorder of the brain and not simply a dysfunction of the "mind."27 Roger N. Rosenberg, MD, and Kenneth Z. Altshuler, MD, provided personal encouragement and administrative support. The neurology house staff, especially Robert Brown, MD, and Donald Orr, MD, provided under¬ standing care of our patients. Richard Parker, PhD, performed the serum cortisol measurements. Jo Ann Brown and Carol Fairchild, MSN, assisted in data compilation. Denise Wernet and Marie Marks provided secretarial assistance. References 1. Flor-Henry P, Koles ZJ: EEG studies in depression, mania and normals: Evidence for partial shifts of laterality in the affective psychoses. Adv Biol Psychiatry 1980;4:21-43. 2. Stevens JR, Bigelow L, Denney D, et al: Telemetered EEG-EOG during psychotic behaviors of schizophrenia. Arch Gen Psychiatry 1979;36:251\x=req-\ 262. Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016 3. Hommes OR, Panhuysen LH: Depression and cerebral dominance: A study of bilateral intracarotid Amytal in 11 depressed patients. Psychiatr Neurol Neurochir 1971;74:259-270. 4. Flor-Henry P: On certain aspects of the localization of the cerebral systems regulating and determining emotions. Biol Psychiatry 1979;14:677\x=req-\ 698. 5. Gruzelier J, Venables P: Bimodality and lateral asymmetry of skin conductance orienting actuity in schizophrenics: Replication and evidence of lateral asymmetry in patients with depression and disorders of personality. Biol Psychiatry 1974;8:55-73. 6. Myslobodsky MS, Horesch N: Bilateral electrodermal activity in depressive patients. Biol Psychiatry 1978;6:111-120. 7. Schweitzer L, Becker E, Welsh H: Abnormalities of cerebral lateralization in schizophrenic patients. Arch Gen Psychiatry 1978;35:982-985. 8. Wexler BE, Heninger GR: Alterations in cerebral laterality during acute psychotic illness. Arch Gen Psychiatry 1979;36:278-284. 9. Gur RE: Left hemisphere dysfunction and left hemisphere overactivation in schizophrenia. J Abnorm Psychol 1978;87:226-238. 10. Gur RE: Motoric laterality imbalance in schizophrenia: A possible concomitant of left hemisphere dysfunction. Arch Gen Psychiatry 1977; 34:33-37. 11. Alpert M, Rubinsten H, Kesselman M: Asymmetry of information processing in hallucinators and nonhallucinators. J Nerv Ment Dis 1976;162:258-265. 12. Wexler BE: Cerebral laterality and psychiatry: A review of the literature. Am J Psychiatry 1980;137:279-291. 13. Gruzelier J, Flor-Henry P (eds): Hemispheric Asymmetries of Functions in Psychopathology. New York, Elsevier North Holland Inc, 1979. 14. Heilman KM, Scholes R, Watson RT: Auditory affective agnosia. Neurol Neurosurg Psychiatry 1975;38:69-72. 15. Tucker DM, Watson RT, Heilman KM: Discrimination and evocation of affectively intoned speech in patients with right parietal disease. J Neurology 1977;27:947-950. 16. Ross ED, Mesulam M-M: Dominant language functions of the right hemisphere? Prosody and emotional gesturing. Arch Neurol 1979;36:144\x=req-\ 148. 17. Ross ED: The aprosodias: Functional-anatomical organization of the affective components of language in the right hemisphere. Arch Neurol 1981;38:561-569. 18. Ross ED, Harney JH, deLacoste-Utamsing C, et al: How the brain integrates affective and propositional language into a unified brain function: Hypothesis based on clinicopathological correlations. Arch Neurol, 1981;38:745-748. 19. Bear D, Fedio P: Quantitative analysis of interictal behavior in temporal lobe epilepsy. Arch Neurol 1977;34:454-467. 20. Gainotti G: Emotional behavior and hemisphere side of lesion. Cortex 1972;8:41-55. 21. Rossi GF, Rosadini G: Experimental analysis of cerebral dominance in man, in Millikan CH, Darley FL (eds): Brain Mechanisms Underlying Speech and Language. New York, Grune & Stratton Inc, 1967, pp 167-174. 22. Dimond SJ, Farrington L, Johnson P: Differing emotional responses from right and left hemispheres. Nature 1976;261:690-692. 23. Schwartz GE, Davidson RJ, Maer F: Right hemisphere lateralization for emotion in the human brain: Interactions with cognition. Science 1975;190:286-288. 24. Sackeim HA, Gur RC: Lateral asymmetry in intensity of emotional expression. Neurophychologia 1978;16:473-481. 25. Jackson JH: On affections of speech from disease of the brain. Brain 1915;38:106-174. 26. Benson DF: Aphasia, Alexia, and Agraphia. New York, Churchill Livingstone, 1979. 27. Adams F (ed-trans): The Genuine Works of Hippocrates. Baltimore, Williams & Wilkins Co, 1939. 28. Spitzer RL, Endicott J, Robins E: Research Diagnostic Criteria: Rationale and reliability. Arch Gen Psychiatry 1978;35:773-782. 29. Feighner JP, Robins E, Guze SB, et al: Diagnostic criteria for use in psychiatric research. Arch Gen Psychiatry 1972;26:57-63. 30. American Psychiatric Association, Committee on Nomenclature and Statistics: Diagnostic and Statistical Manual of Mental Disorders, ed 3. Washington, DC, American Psychiatric Association, 1980. 31. Folstein MF, Maiberger R, McHugh P: Mood disorder as a specific complication of stroke. J Neurol Neurosurg Psychiatry 1977;40:1018-1020. 32. Carroll BJ, Curtis GC, Mendels J: Neuroendocrine regulation in depression: I. Limbic system-adrenocortical dysfunction. Arch Gen Psychia- try 1976;33:1039-1044. 33. Carroll BJ, Curtis GC, Mendels J: Neuroendocrine regulation in depression: II. Discrimination of depressed from nondepressed patients. Arch Gen Psychiatry 1976;33:1051-1058. 34. Brown WA, Johnston R, Mayfield D: The 24-hour dexamethasone suppression test in a clinical setting: Relationship to diagnosis, symptoms and response to treatment. Am J Psychiatry 1979;136:543-547. 35. Schlesser MA, Winokur G, Sherman BM: Hypothalamic-pituitaryadrenal axis activity in depressive illness: Its relationship to classification. Arch Gen Psychiatry 1980;37:737-743. 36. Carroll BJ, Feinberg M, Greden JF, et al: A specific laboratory test for the diagnosis of melancholia: Standardization, validation, and clinical utility. Arch Gen Psychiatry 1981;38:15-22. 37. Nelson JC, Charney DS: The symptoms of major depressive illness. Am J Psychiatry 1981;138:1-13. 38. Klein DF: Endogenomorphic depression. Arch Gen Psychiatry 1974; 31:447-454. 39. Rush AJ, Giles DE, Roffwarg HP, et al: Sleep EEG and dexamethasone suppression test findings in unipolar major depressive disorder. Biol Psychiatry, to be published. 40. Rush AJ, Beck AT, Kovacs M, et al: Comparative efficacy of cognitive therapy and pharmacotherapy in the treatment of depressed outpatients. Cog Ther Res 1977;1:17-37. 41. Kovacs M, Rush AJ, Beck AT, et al: Depressed outpatients treated with cognitive therapy or pharmacotherapy: A one-year follow-up. Arch Gen Psychiatry 1981;38:33-39. 42. Weissman MM: The psychological treatment of depression: Evidence for the efficacy of psychotherapy alone, in comparison with, and in combination with pharmacotherapy. Arch Gen Psychiatry 1979;36:1261\x=req-\ 1269. 43. Gillespe RD: The clinical differentiation of types of depression. Guy Hosp Rep 1929;79:306-344. 44. Benson DF: Psychiatric aspects of aphasia. Br J Psychiatry 1973; 123:555-566. 45. Robinson RG, Szetela B: Mood change following left hemispheric brain injury. Ann Neurol 1981;9:447-453. 46. Hamilton M: A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23:56-62. 47. Wilson SAK: Pathological laughing and crying. JNeurol Psychopathol 1924;4:299-333. 48. Benson DF, Geschwind N: Psychiatric conditions associated with focal lesions of the central nervous system, in Arietti S, Reiser N (eds): American Handbook of Psychiatry. New York, Basic Books Inc Publishers, 1975, vol 4, pp 208-243. 49. Gazzaniga MS, Le Doux JE: The Integrated Mind. New York, Plenum Press Inc, 1978. 50. Kaada BR: Stimulation and regional ablation of the amygdaloid complex with reference to functional representations, in Eleftheriou BE (ed): The Neurobiology of the Amygdala. New York, Plenum Press Inc, 1972, pp 205-281. 51. Gloor P: Temporal lobe epilepsy: Its possible contribution to the understanding of the functional significance of the amygdala and of its interaction with neocortical-temporal mechanisms, in Eleftheriou BE (ed): The Neurobiology of the Amygdala. New York, Plenum Press Inc, 1972, pp 423-457. 52. Zolovick AJ: Effects of lesions and electrical stimulation of the amygdala on hypothalamic-hypophyseal regulation, in Eleftheriou BE (ed): The Neurobiology of the Amygdala. New York, Plenum Press Inc, 1972, pp 643-683. 53. Adams GF, Hurwitz LJ: Mental barriers to recovery from strokes. Lancet 1963;2:533-537. 54. Bell DS: Psychiatric aspects of cerebral vascular disease. Med J Aust 1966;2:829-833. 55. Benson DF, Geschwind NG: The aphasias and related disturbances, in Baker AB, Baker LH (eds): Clinical Neurology. Hagerstown Md, Harper & Row Publishers Inc, 1977, vol 1, pp 10-11. 56. Dimond SJ: Brain circuits for consciousness. Brain Behav Evol 1976; 13:376-395. 57. Dimond SJ: Disconnection and psychopathology, in Gruzelier J, Flor-Henry P (eds): Hemisphere Asymmetries of Function in Psychopathology. New York, Elsevier North Holland Inc, 1979, pp 35-46. 58. DeLacoste-Utamsing C, Holloway RL, Kirkpatrick JL, et al: Anatomical and quantitative aspects of the human corpus callosum. Soc Neurosci Abstr 1981;7:382. 59. Flor-Henry P: The endogenous psychoses: A reflection of lateralized dysfunction of the anterior limbic system, in Livingston KE, Hornykiewicz O (eds): Limbic Mechanisms. New York, Plenum Press Inc, 1978, pp 389-404. 60. Farkas T: Symposium Session, Biochemical imaging of the brain: Clinical discussion. Read before the 35th annual convention and scientific program of the Society of Biological Psychiatry, Boston, Sept 5-7, 1980. 61. Mathew RJ, Meyer JS, Francis DJ, et al: Cerebral blood flow in depression. Am J Psychiatry 1980;137:1449-1450. 62. Matsui T, Hirano A: An Atlas of the Human Brain for Computerized Tomography. Tokyo, Igaku-Shoin, 1978. Downloaded From: http://archpsyc.jamanetwork.com/ by a New York University User on 02/21/2016