Daniel B. Hier

The Stroke Data Bank: design, methods, and baseline characteristics.

The National Institute of Neurological and Communicative Disorders and Stroke initiated the Stroke Data Bank, which is a multicenter project to prospectively collect data on the clinical course and sequelae of stroke. Additional objectives were to provide information that would enable a standard diagnostic clinical evaluation, to identify prognostic factors, and to provide planning data for future studies. A brief description of the structure and methods precede the baseline characterization of 1,805 patients enrolled in the Stroke Data Bank between July 1983 and June 1986. Two thirds of these patients were admitted within 24 hours after stroke onset. Medical history, neurologic history, and hospitalization summaries are presented separately for the following stroke subtypes: infarction, unknown cause; embolism from cardiac source; infarction due to atherosclerosis; lacune; parenchymatous or intracerebral hemorrhage; subarachnoid hemorrhage; and other. The utility and limitations of these data are discussed.

Morning increase in onset of ischemic stroke.

The time of onset of ischemic stroke was determined for 1,167 of 1,273 patients during the collection of data by four academic hospital centers between June 30, 1983, and June 30, 1986. More strokes occurred in awake patients from 10:00 AM to noon than during any other 2-hour interval. The incidence of stroke onset declined steadily during the remainder of the day and early evening. The onset of stroke is least likely to occur in the late evening, before midnight.

Recovery of behavioral abnormalities after right hemisphere stroke

We studied recovery of function in 41 patients with right hemisphere stroke. Recovery was rapid for left neglect, prosopagnosia, anosognosia. and unilateral spatial neglect on drawing (USN). Recovery was slower for hcmianopia. hemiparesis, motor impersistence, and extinction. Rates of recovery were intermediate for constructional apraxia and dressing apraxia. Sex had no influence on the rate of recovery. Younger patients recovered from prosopagnosia more rapidly than older patients. Patients with smaller lesions recovered more quickly from anosognosia, USN, and hemiparcsis than patients with larger lesions. Patients with hemorrhages recovered more rapidly from constructional apraxia, neglect, and motor impersistence than patients with infarcts. Recovery of function and the factors influencing recovery can by studied systematically by life table methods.

Stroke recurrence within 2 years after ischemic infarction.

We prospectively studied stroke recurrence in 1,273 patients with ischemic stroke who were entered into the Stroke Data Bank. Median follow-up was 13 months. The 2-year cumulative recurrence rate among these patients was 14.1%. Age, sex, race, history of hypertension, atrial fibrillation, or transient ischemic attacks, and stroke location were not associated with a higher risk of stroke recurrence. Patients with an elevated blood pressure, an abnormal initial computed tomogram, or a history of diabetes mellitus were at a higher risk of stroke recurrence. In contrast, patients with an infarct of unknown cause were at a lower risk of stroke recurrence than patients with a defined stroke mechanism, such as lacune, embolism, or atherosclerosis. A multivariate model suggests that patients at the lowest risk for stroke recurrence have a low diastolic blood pressure, no history of stroke, no history of diabetes mellitus, and an infarct of unknown cause.

Behavioral abnormalities after right hemisphere stroke

We evaluated 41 patients with unilateral right hemisphere strokes for hemiparesis, hemianopia, constructional apraxia, neglect, extinction, unilateral spatial neglect on drawing (USND), dressing apraxia, anosognosia, prosopagnosia, and motor impersistence. Low correlations were found among USND, neglect, and extinction. Constructional apraxia showed a higher correlation with USND than with hemianopia. Motor impersistence and anosognosia correlated with the severity of the hemiplegia. Left neglect, motor impersistence, and anosognosia tended to occur only with large strokes. Injury to the right parietal lobe appears to be an important determinant of USND and constructional apraxia. In most cases of motor impersistence, left neglect, and anosognosia, there was injury to the right parietal lobe and also to structures beyond the parietal lobe. In two cases, small deep lesions produced behavioral abnormalities comparable to those of larger superficial cortical lesions.

Language disintegration in dementia: Effects of etiology and severity ☆

The speech characteristics on a standardized picture description task of 26 subjects with presumed senile dementia of the Alzheimer type (SDAT) and 13 subjects with stroke-related dementia (SRD) were compared to 15 normal subjects over age 59 years. Compared to the normal subjects, the dementia subjects used fewer total words, fewer unique words, fewer prepositional phrases, fewer subordinate clauses, and more incomplete sentence fragments. Lexical deficits tended to be more severe than syntactic ones, confirming prior suggestions that lexicon is more vulnerable to disruption in dementia than syntax. Greater dementia severity among the SDAT subjects was associated with marked difficulties in accessing the mental lexicon (increased use of empty words, indefinite anaphora, and pronouns). Greater dementia severity in the SRD subjects was associated with laconic speech that was syntactially less complex. Diffuse brain injury (as typified by SDAT) appears to disproportionately affect lexicon whereas multifocal injury (as typified by SRD) has a disproportionate effect on syntax (assuming that focal lesions of the posterior language zone have been excluded). The speech characteristics of the mild SDAT subjects showed similarities to those of anomic or semantic aphasia whereas the speech of the more advanced SDAT subjects showed similarities to Wernicke aphasia or transcortical sensory aphasia. The speech of the subjects with more severe SRD showed some similarities to Broca aphasia. The most important nonlinguistic deficit in both the SRD and the SDAT groups was a failure to make relevant observations during the picture description task. Perseverations were present in the speech of both the SRD and SDAT subjects, whereas aposiopesis, logorrhea, and palilalia were more typical of the SDAT subjects. Laconic speech was more characteristic of the SRD subjects.

Auditory middle latency responses (MLRs) in patients with cortical lesions

Auditory middle latency response (MLRs) and auditory brain stem responses (ABRs) were simultaneously recorded in 24 patients with cortical lesions primarily affecting the temporal lobes. Site of lesion was documented by computerized tomography (CT) scan and behavioral profiles assessing language and other higher cortical functions were obtained. In patients with normal ABRs and either left or right hemisphere lesions, MLR components Na and Pa obtained at the vertex were of normal shape and latency. Exceptions to this occurred in 2 patients: one with bilateral temporal lobe lesions, the second with an infraventricular left temporal lobe lesion extending into the thalamic radiations. Although Na and Pa shape and latency were for the most part unaltered, Pa amplitude tended to cluster at the low end and below normal values. MLR recorded in the coronal plate showed Pa amplitude to be attenuated or absent over the damaged temporal lobe relative to the vertex or the intact hemisphere. This finding contrasts with data from normal subjects where Pa amplitude is largest at the vertex and essentially symmetrical about the temporal lobes. Patients showing an atypical amplitude distribution tended to have lesions involving auditory cortex and adjacent white matter projections. No obvious correlations between MLR abnormalities and behavioral findings regarding receptive and expressive language processes were found. Pa appears to be affected by temporal lobe lesions involving auditory cortex and thalamic projections. Our findings support the hypothesis that Pa is bilaterally generated by two symmetrical, vertically oriented dipole sources located about the temporal lobes.

Racial differences in the distribution of anterior circulation occlusive disease

We compared clinical and angiographic features of 26 white and 45 black patients with symptomatic occlusive cerebrovascular disease. White patients had more transient ischemic attacks, carotid bruits, and more severe occlusive disease of the internal carotid artery origin. Blacks had more severe disease of the middle cerebral artery stem and supraclinoid internal carotid arteries. Differences were not explained by racial differences in the prevalence of hypertension, diabetes, hypercholesterolemia, or ischemic heart disease. Since the middle cerebral artery lesions in blacks do not correlate with other accepted epidemiologic, clinical, and laboratory markers of atherosclerosis, the lesions may arise from a disorder that differs from atherosclerosis.

Autism and Unfavorable Left-Right Asymmetries of the Brain.

Utilizing computerized brain tomography, left-right morphologic asymmetries of the parietooccipital region were judged in 16 autistic patients, 44 mentally retarded patients, and 100 miscellaneous neurological patients. In 57% of the autistic patients the right parietooccipital region was wider than the left, while this pattern of cerebral asymmetry was found in only 23% of the mentally retarded patients and 25% of the neurological patients. It is suggested that unfavorable morphologic asymmetries of the brain near the posterior language zone may contribute to the difficulties autistic children experience in acquiring language.

Headache in acute cerebrovascular disease

Headache features were compared in 51 patients with acute subarachnoid hemorrhage (SAH), 61 with intraparenchymal hemorrhage (IPH), and 160 with ischemic stroke (IS). SAH patients had more sentinel headaches, more onset headaches, and more bilateral and severe onset headaches than patients with IPH or IS. Vomiting with onset headache was more common in SAH and IPH. In stepwise logistic regression analysis, onset headache and vomiting were direct predictors of SAH, but were inversely related to IS. Sentinel headache was not a predictor of underlying stroke mechanism. The data suggest that some headache features are more frequently associated with particular stroke subtypes and that onset headache and vomiting may be important indicators of stroke mechanism.