1000 STROKE Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 8. Manuck SB, Kaplan JR, Clarkson TB: Behaviorally induced heart rate reactivity and atherosclerosis in cynomolgus monkeys. Psychosom Med 45: 95-108, 1983 9. Kaplan JR, Manuck SB, Clarkson TB, et al. Social stress and atherosclerosis in normocholesterolemic monkeys. Science 220: 733-735, 1983 10. Lipid Research Clinics Program. Manual of Laboratory Operations: Lipid and lipoprotein analysis. DHEW publication no. (NIH) 76-628. Bethesda, Maryland: National Institutes of Health, 1974 11. Williams RB: Psychophysical differences between Type A and B individuals that may lead to coronary heart disease. In: Dembroski TM (ed). Proceedings of the Fonim on Coronary Prone Behavior. DHEW publication no. (NIH) 78-1451:185-191, 1977 12. Siegel J: Nonparametric statistics. New York, McGraw Hill, 1956 13. Castelli WP: High density lipoproteins: An overview. In Gotto AM Jr, Smith LC, Allen B (eds.), Atherosclerosis V. New York, Springer-Verlag, 478-483, 1980 14. Kirk RE: Experimental design (2nd Edition), Belmont, CA, Brooks/Cole, 1982 15. Solberg LA, McGarry PA, Moossy J, et al: Distribution of cerebral atherosclerosis by geographic location, race and sex. Lab Invest 18: 604-612, 1968 VOL 15, N o 6, NOVEMBER-DECEMBER 1984 16. Kuller L, Reisler DM: An explanation for variations in distribution of stroke and arteriosclerotic heart disease among populations and racial groups. Am J Epidemiol 93: 1-9, 1971 17. Kuller LH: Epidemiology of stroke. In: Schoenberg BS (ed), Advances in Neurology, Vol. 19. New York, Raven Press, 281-311, 1978 18. Lavy S: Medical risk factors for stroke. In: Goldstein M et al (eds), Advances in Neurology, Vol. 25. New York, Raven Press, 127— 133, 1979 19. Wolf PA, Kannel WB, Dawber TR: Prospective investigations: The Framingham study and the epidemiology of stroke. In: Schoenberg BS (ed), Advances in Neurology, Vol 19. New York, Raven Press, 107-120, 1978 20. Corbett WT, Schey HM, Lehner NDM, et al: Standardized method for recording blood pressure in anesthetized Macaca fascicularis. Lab Anim 15: 37-40, 1981 21. Hollander W, Geer JC, Garcia JH: Aggravation of cerebral atherosclerosis by hypertension. Arteriosclerosis 3: 479a, 1983 22. Glagov S: Low heart rate retards coronary atherosclerosis. Arteriosclerosis 3: 466a, 1983 23. Hamm TE, Kaplan JR, Clarkson TB, et al: Gender and social factors in the atherosclerosis of cynomolgus monkeys. Atherosclerosis 48: 221-233, 1983 CT-CBF Correlations of Cognitive Deficits in Multi-Infarct Dementia Y A S U H I S A K I T A G A W A , M . D . , P H . D . , JOHN S. MEYER, M . D . , HISAO TACHIBANA, M . D . , P H . D . K A R L F. MORTEL, P H . D . , AND ROBERT L. ROGERS, M . A . SUMMARY Fifteen right-handed patients with Multi-Infarct Dementia underwent cognitive testing by the Jacobs Mini-Mental Scale (MMQ), and xenon contrast CT scanning. Local cerebral blood flow (LCBF) and local partition coefficient (LX) values were measured by stable xenon contrast CT scanning and potential methodological errors were discussed. Reduced values were graded: 0 = normal, 1 = mild, 2 = moderate, 3 = severe. Graded values were pooled and plotted on composite brain maps to display locations of abnormal LX and LCBF values. Topographic brain maps, showing most frequent locations of reduced L>. values, confirmed the common anatomical locations of multiple cerebral infarcts to be distributed in both thalami, temporal lobes, basal ganglia, left internal capsule and right cingulate cortex. Gray matter flow values were reduced in similar cortical and subcortical regions. There were no correlations between MMQ scores and reduced LCBF values for caudate and lenticular nuclei. Direct and statistically significant correlations were found between reduced MMQ scores and mean LCBF values for left or right frontal cortex, left or right temporal cortex and left or right thalamus. Subgrouping MMQ tests according to functions assessed, indicated that left mid-temporal ischemia correlated with dyscalculia and memory disturbances while ischemia of both frontal lobes correlated with disorientation to time and place. Stroke Vol 15, No 6, 1984 DEMENTIA is becoming a serious public health problem in our aging population. Epidemiological studies have shown that dementia is present in 4.4-8% of populations over age 65 years.'^ Pathological investigations have revealed that of the commonest forms of dementia, about 50% of elderly demented patients sufFrom the Cerebrovascular Research Laboratories, VA Medical Center, Department of Neurology, Baylor College of Medicine; and Department of Psychology, University of Houston, Houston, Texas. This work was supported, in part, by the Veterans Administration Medical Center and by a Grant provided by Mr. Harry K. Smith and Mr. Albert K. Smith, Houston, Texas. Address correspondence to: John S. Meyer, M.D., Director, Cerebrovascular Research Laboratories, VA Medical Center, 2002 Holcombe Boulevard, Houston, Texas 77211. Received October 17, 1983; revision #2 accepted May 11, 1984. fer from senile dementia of Alzheimer's type (SDAT), 20% from multi-infarct dementia (MID) and 10% have both diseases (MIX).1 Differentiation between SDAT and MID is a commonly encountered diagnostic problem in the clinic. Hachinski's ischemic score5 has proven clinically useful in distinguishing between these most common types of dementia and the validity of this scoring system has been neuropathologically established in a series of autopsied cases by Rosen, et al.6 The present communication will attempt to correlate cognitive deficits measured in patients with MID with quantitative measurements of local cerebral blood flow measured by xenon contrast CT scanning. In the majority of cases, MID appears to be a cere- 1001 STABLE XENON, CTCBF IN MID/Kitagawa et al Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 bral manifestation of long-standing hypertension. At autopsy hyperplastic cerebral arteriosclerosis is evident with bilateral multiple small lacunar infarcts.7 The severity of the dementia is not only proportional to the cumulative severity of the lacunar state, but also to the locations of the lacunes. 8 From a cognitive point of view, patients with MID may be considered as a useful model. They offer an opportunity for studying those regions of the brain, which if damaged by multiple small infarcts, may give rise to dementia. Cerebral blood flow studies have provided some insights concerning the pathogenesis and differences between MID and SDAT.9"21 It has been known for 30 years that organic dementia is consistently accompanied by bilateral hemispheric reductions of cerebral blood flow and oxygen uptake. 9 The development of regional methods for measuring cerebral blood flow, such as I33Xe clearance methods, made some regional quantitation possible between cognitive deficits and the severity and nature of dementia.10"21 However, these two-dimensional radioisotopic methods ignore or underestimate many zones of zero or reduced flow because clearance of l33Xe from bordering zones obscures them (so-called "look through" phenomenon). There are other problems which impair the resolving powers of the 133Xe method including Compton scatter and lack of knowledge of brain-tissue solubility coefficients (IA). With the advent of computerized reconstruction techniques for transmission and emission tomography it is now possible to image the brain in three dimensions with high resolution. Positron emission tomography provides considerable advantages of measuring both blood flow and metabolism but has the disadvantages of enormous cost, prolonged time consumption for patients and investigators and relatively poor resolution compared to computerized x-ray transmission tomography.22-23 TABLE 1 Case Series Age Case numbers The excellent resolving powers of x-ray transmission tomography, utilizing the CT scanner for measuring both LA. and LCBF during inhalation of stable xenon gas (CTCBF method), overcomes the limitations of the 133Xe clearance method and provides the best resolution of any method presently available for measuring cerebral blood flow (circa 80 cu mm).24"32 The purpose of the present communication is to report regional patterns of reduced LA. and LCBF values measured by the CTCBF method in 15 patients with well-established MID and to correlate anatomical zones of regional ischemia with cognitive deficits measured within a few days of the LCBF and LA. determinations. Case Series Fifteen right-handed and cooperative patients with well-defined MID were selected from a larger series with this disease being followed in this laboratory. The diagnosis in all patients was well-established. It was confirmed by history of sudden onset and fluctuating course, the presence of risk factors for cerebrovascular disease and by medical and neurological examinations showing hypertension, cerebral arteriosclerosis by neurovascular examination (bruits, hypertensive retinal changes, etc.) and focal neurological signs. The cognitive deficits were confirmed by the Mini-Mental Status examination (MMQ) of Jacobs, et al33 modified after that described by Folstein, et al.34 Scores in normal age-matched volunteers are between 29-31 but the patient's scores were between 11-20 (table 1). The patient's ages ranged from 57 to 85 years with a mean age of 71.5 ± 10.1. All patients had a history of chronic but not malignant or accelerated hypertension, recurrent attacks of cerebral ischemia including transient ischemic attacks, reversible ischemic neurological deficits and small strokes. Patients with large or aphasic strokes were excluded. All showed evidence MABP (mmHg) (in years) Gender 1 84 F 110 2 72 F 104 3 54 M 129 4 65 M 97 5 81 M 98 6 72 M 123 7 79 M 110 8 75 M 97 9 85 F 100 10 60 F 109 11 57 M 99 12 74 F 93 13 79 F 97 14 77 F 93 15 59 F 89 71.5±10.1 7M.8F Mean values 103.2± 11.2 PECOj (mmHg) 34.6 38.5 34.2 35.6 37.1 34.2 35.7 35.9 28.7 35.9 35.7 33.0 40.2 34.2 35.7 35.3±2.6 Ischemic score MMQ 11 15 7 13 value 9 17 9 11 8 18 11 20 7 19 12 13 7 16 10 18 9 19 10 14 9 19 7 14 7 8.9±1.7 16 16.1 ±2.7 1002 STROKE Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 of arteriosclerosis elsewhere in the body such as arteriosclerotic heart disease, arteriosclerotic retinopathy and/or bruits over the carotid and vertebral arteries. There were focal neurological signs on examination, a history of focal neurologic symptoms, the presence of at least one or more risk factors for atherothrombotic stroke, step-wise neurologic deterioration, and multisegmental cerebral atherosclerosis revealed in the majority of cases by aorto-cranial arteriogTaphy. CT scanning in all cases showed evidence of multiple cerebral infarcts (low density lesions, asymmetrical or focal cortical atrophy, ventricular asymmetry). Cases with recent stroke (within 30 days of the last stroke) were excluded to avoid any influences ofreactivehyperemia upon CBF. The ischemic index criteria of MayerGross, as described by Hachinski, et al,14 were part of the evaluation used for differentiating MID from SDAT. Patients with an ischemic index of 7 or more were considered qualified to meet the rigid criteria required for establishing an unequivocal clinical diagnosis of MID. Patients with Bingwanger's subcortical leukoencephalopathy were excluded by the criteria listed above, plus the characteristic CT findings of a diffuse rather than multifocal low density lesions by CT examination.35 Table 1 displays the mean age, gender, mean arterial blood pressure, PECO2, Hachinski's Ischemic Score, and MMQ at the time of measurement of LCBF and LX values. Methods Measurements of LCBF and LX were carried out utilizing the stable xenon CTCBF method for each region of interest by means of a single compartmental analysis program. Details of the method have been reported in previous publications.24 2J28-31 After explaining the procedure to each patient and family, including a simple explanation of the exposure to x-rays and the discomfort of lying in the scanner for 20 minutes, informed consent* was signed by both patient and the legal guardian. Patients were asked to fast for 6 hours before the CT scanning to avoid the possibility of vomiting during the procedure. Each patient was then placed in the EMI 1010 CT scanner to obtain two back-to-back 8 mm brain slices, 8 mm apart 4.0 cm and 4.8 cm above the orbito-meatal line and in that horizontal plane. The two slices selected provided optimal information about gray and white matter of frontal, temporal, parietal and occipital lobes, internal capsule, thalamus and basal ganglia. To avoid excessive radiation exposure only two slices were measured for LCBF and LX values. The brain stem, cerebellum and upper hemispheric levels were not measured. Body nitrogen was displaced by inhalation of 100% oxygen for 10 minutes piror to the CBF measurements. During denitrogenation, 3-4 serial non-contrast scans were made. Three *The informed consent form used and the protocol for measuring CBF by the CT xenon contrast method were approved by the Institutional Review Board, Veterans Administration Medical Center, Houston. VOL 15, No 6, NOVEMBER-DECEMBER 1984 to four xenon contrast scans were obtained between the 2nd and 9th minute of inhalation of 35-37.5% stable xenon gas mixed with 62.5-65% oxygen. Oxygen in these concentrations is a mild vasoconstrictor, reducing LCBF values for gray matter by at most 5% and white matter by considerably less.23 These minimal vasoconstrictive effects are within the experimental error of the method which is ± 5%. The xenon mixture was administered through a semi-closed partial rebreathing system. The rate of increase in alveolar concentration of xenon gas was carefully controlled so that gradual increases in alveolar xenon concentration resulted. Dupont Tedlar gas sampling bags (Model 130a, AeroVironment Inc., Pasadena, California 91107) were used to deliver the gas mixtures. These are made of Tedlar plastic which is impermeable to xenon gas. This has proved to be cost-effective, since the unused portion of the gas remains in the bag and can be used for later patients. A gradual build-up of alveolar xenon concentrations provides optimal curve fitting and minimizes any subanesthetic effects of xenon gas. Local CT enhancements of the brain during build-up by xenon were measured in selected regions of interest from small volumes (80-250 cu mm3) located in homogeneous gray or white matter. End-tidal xenon gas concentrations were monitored with a Gow-Mac thermoconductivity analyzer. Inspiratory and expiratory gas concentrations were recorded on a polygraph. Since end-tidal and arterial gas concentration are in equilibrium, these values were converted to A H changes for blood using a proportionality constant derived from Kelcz's formula36 and corrected for the patient's hematocrit. The time consumed while making these measurements is 15-20 minutes at cost of $50 per patient. LCBF and LX values were calculated by means of a computer program utilizing least squarefittingof saturation curves for both end-tidal and brain tissue A H values fitted to infinity. Blood pressure, PECO2, PEO2, EKG and EEG were concurrently recorded on the polygraph. Local CT enhancements of the brain by stable xenon, during inhalation of the gas, were measured as changes in Hounsfield units plotted from a baseline of 3-4 non-contrast CT scans and the 3 or more data points measured by the CT scanner during saturation with the contrast agent. Saturation curves were then plotted prior to computer analysis. These were considered satisfactory if they appeared curvilinear and showed deviations of ± 0.5 Hounsfield units or less derived for 3 or more points. If these criteria were not met that particular curve was rejected as unsatisfactory and was not analyzed by the computer. Such rejections occurred in about 20% of all measurements and were evenly distributed among controls and patients. The smoothing of the curves, necessary because of the low signal to noise ratio is a potential source of error of the method which could influence LX values in particular and therefore LCBF values to a lesser degree. Such errors could result in misclassifications of STABLE XENON, CTCBF IN MID/Kitagawa et al TABLE 2 Methods Used on Brain Maps for Classifying Reduced Local Blood Flow and Lambda Values White matter Gray matter LCBF values 60.1->60.1 60-58 57.9-55 54.9-<54.9 LX values O.8O1->O.8O1 0.80-0.78 0.779-0.75 0.749-<0.749 Points assigned 0 1 2 3 0 1 2 3 LCBF values 20.1->20.1 20-18 17.9-15 14.9-<14.9 LX values 1.301->1.301 1.30-1.28 1.279-1.25 1.249-< 1.249 Points assigned 0 1 2 3 0 1 2 3 Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 the data by the methods to be described later when discussing tables 2 and 3. The volumes or voxels of interest selected for each measurement measured between 80-250 cu mm derived from what appeared by their appearance on plain CT scanning, to be composed of homogeneous gray or white matter. If when sampling gray matter, some white matter was included by error due to partial volume effects, this became apparent from calculated LX values, which showed values higher than for normal gray matter (0.86 ± 0.04). If LA's for gray matter were excessive this sample was rejected as inhomogeneous, and another voxel was sampled. In practice, this was rarely a problem, particularly when selecting homogeneous volumes of white matter, since white matter of the human brain is usually thicker than corti- cal gray matter. In other words, for anatomical reasons partial volume effects were rare when sampling white matter and could be corrected when sampling gray. The possibility of errors associated with CTCBF measurements has been reviewed in detail elsewhere.28' "• 38 In order to plot the most common tomographic localizations for severe reductions of LCBF and LX. values measured among the entire series of patients, reduced LCBF and LX values were classified into grades of severity between 0, 1, 2, and 3 (table 2). Regions with flow values below 10 ml/100 g brain/min were seldom encountered and were not included in order to avoid diffusion errors when measuring LX, values. Total scores for reduced flows and X values for representative regions of the brain were then plotted on composite brain maps as pooled data for all 15 cases. For example, 3 points were assigned to regions showing severe reductions of LCBF and LX values, 2 points were assigned for moderate reductions and 1 point for mild and zero points for regions having normal LCBF and LX values. In this age group, normal values for gray matter LCBF were considered to be 60 ml/lOOg brain/min or above, as judged from similar measurements in age-matched normal volunteers.28 Values below 55 ml/lOOg brain/min were considered to be severely reduced. Normal values for white matter LCBF were considered to be values above 20 ml/lOOg brain/min, while those between 10-15 ml/lOOg brain/ min were considered to be severely reduced. Regarding LX values, gray matter values below 0.80 were considered to be reduced and those below 0.75 were considered severely reduced. White matter LX values were considered abnormal if below 1.30 and severely TABLE 3 Correlation of LCBF Values with MMQ Scores. Correlation Coefficients were Significant between MMQ Scores and Mean LCBF Values for Left or Right Frontal Cortex, Left or Right Temporal Cortex and Left or Right Thalamus Regression Correlation Statistical line coefficient (r) significance N* y = 0.80x + 46.9 15 0.79 P < 0.01 0.52 y = 0.87x + 46.9 15 P < 0.05 y = 1.06x x 42.5 0.59 15 P < 0.05 N.S. N.S. 0.10 7 0.32 12 N.S. N.S. 0.74 y = O.78x + 47.6 12 P < 0.01 y = 1.61x + 32.7 0.58 15 P < 0.05 0.21 N.S. N.S. 6 11 N.S. N.S. 0.06 N.S. 0.36 15 N.S. N.S. 0.50 13 N.S. 0.62 11 y = 0.87x + 45.4 < P 0.05 N.S. 0.44 9 N.S. 12 N.S. 0.19 N.S. 11 y = 0.80x + 44.9 0.61 < P 0.05 7 N.S. 0.01 N.S. 15 N.S. N.S. 0.09 •Number ofregionsof interest examined. Someregionswere excluded for technical reasons, due to head movement or poor head positioniong. Region of interest Cortical gray matter Lt. frontal matter Lt. temporal cortex Lt. parietal cortex Lt. occipital cortex Rt. frontal cortex Rt. temporal cortex Rt. parietal cortex Rt. occipital cortex Subcortical gray matter Lt. caudate nucleus Lt. thalamus Lt. putamen Rt. caudate nucleus Rt. thalamus Rt. putamen White matter 1003 1004 STROKE Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 reduced if below 1.25. By pooling the scores for regional reductions, two maps were constructed, one for reduced LA topography and another for reductions of LCBF. Cognitive deficits were quantitatively assessed by the modification described by Jacobs et al of the Folstein "mini-mental state (MMQ)." This was administered immediately before the CBF test in the majority of cases and within one week in the remainder. The validity of this rapid assessment of cognitive mental deficits has been established in a large series of patients. 3334 They correlate well with Wechsler Adult Intelligence Scale (WAIS). A score between 11-20 indicates moderate to moderately severe cognitive impairment but with sufficient insight to understand the test procedure and sign informed consent. They have considerable advantages over the longer psychological test batteries. The latter are time-consuming and patients with dementia tend to tire or refuse to cooperate. The MMQ is designed specifically for organic mental syndromes and takes only 30 minutes or less to complete. As evaluated in this laboratory the weighting of the MMQ scale is as follows: Orientation 5 questions, Attention and Calculation 11 questions, Opposites and Similarities 5 questions, Memory Registration 6 questions and Memory Consolidation 4 questions. Normal age-matched volunteers in this laboratory score 29-31 out of a possible 31. Methods For Data Analysis Bivariate correlation coefficients were calculated between MMQ scores and quantitative LCBF values for each region using the Pearson Product-Moment analysis. The slope and intercept of the bivariate regression line were also calculated in order to compare magnitude of effect. Subsequently, a set of correlation coefficients were also determined between LCBF values and subtests of the MMQ. These included orientation, digit repeating and reversing, calculation, similarities and tests of memory. Significance of correlations were tested against the t-distribution and statistical significance was set at p < .05. Regarding the statistical methods used for analyzing data obtained utilizing the available EMI 1010 scanner in this hospital, the mean standard deviation for human VOL 15, No 6, NOVEMBER-DECEMBER 1984 brain slices measured as Hounsfield units during 4 serial control scans measured in gray or white matter prior to the xenon gas inhalation was ±0.7 Hounsfield units for regions of interest 80 cu mm or larger. During xenon inhalation of mixtures of 35-37.5% xenon the measured A H increases were between 5.5 to 8.5 H.U., provided there was no movement of the head. LA and LCBF values were calculated by the use of a single compartment double-integration program and a DEC 10 computer. This was carried out for each one of multiple regions of interest selected from two adjacent brain sections. Two or more representative LCBF and LA values were pooled to provide the data for each of the anatomical regions of interest for patients and normal volunteers listed in tables 2 and 3. This laboratory has accumulated CTCBF measurements among 245 patients and normal volunteers over the past 6 years. These data have been analyzed blindly and at different times. The data were computed and analyzed first among the normal volunteers without knowledge of results in MID, Alzheimer's disease, cases with both (MIX) and dementia of other types such as normal pressure hydrocephalus, etc. Later data from MID patients were computed and analyzed without knowledge of values for age-matched controls. Later data from the two groups were compared. Regions most affected by minor head movements were usually located in frontal and occipital cortex. If such minor head movement occurred, these data points were discarded as already described. However, the same technical problems were found to exist for both patients and normal controls, so that the numbers of discarded data points were similar in both groups and observer bias is considered unlikely. Results Confirmation Of Regional Anatomical Patterns Of Infarction In MID Figure 1 illustrates a composite brain map displaying the topography for pooled local A values among the group of 15 right-handed patients with MID. Moderately to severely reduced LA values are seen in the thalamus bilaterally, both caudate nuclei, both putamenal nuclei, the left, mid- and inferior temporal corti- C O M P O t m MAIN MAP Or POOLID LOCAL LAMBDA VALUU M 1t PUOXT HAHOfD PATHNT* WITH MM.TMWAHCT DCHZHTU TO SHOW UOIONAL AHATOWC MTTDWS OF IMFAUCTION* FIGURE 1. Multiple infarcts as evidenced by decreased Lk values were most commonly distributed in the thalamus bilaterally, both basal ganglia, both temporal cortices plus the left internal capsule and the right cingulate cortex. STABLE XENON, CTCBF IN MID/Kitagawa et al 1005 COMPOSITE MAIN HAP OF POOLED LOCAL C«F VALUES IW 11 RK3HT HANDED PATIENTS WITH HULT1—MFAftCT DEMENTIA TO SHOW MOIONAL ANATOMIC PATTERNS OF FUHCT1OMAL ISCHEMIA UwlMlon FIGURE 2. Reduced LCBF patterns were measured in the thalamus bilaterally, both basal ganglia, both temporal cortices, the left inferior frontal cortex and the left parietal cortex. Severely Reduced Moderately Reduced '• •---I Reduced • Borderline Reduced Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 ces, the right superior and mid-temporal cortex, the right cingulate cortex, the left internal capsule and left frontal white matter. Mildly reduced LX values are also present in the right fronto-temporal white matter. These regions of reduced X values are distributed bilaterally and in a patchy manner typical of lacunar infarcts.31 They confirm the anatomical sites of multiple cerebral infarctions or zones of reversible ischemia which were sometimes (but not always) apparent as low density lesions by plain CT scanning.31 Regional Anatomic Patterns Of Functional Ischemic Impairment Figure 2 displays a composite brain map illustrating the most common sites of reduced LCBF values derived from pooled data for the entire series of 15 patients with MID. Moderately or severely reduced flow patterns are seen in the,fhalamus bilaterally, both caudate nuclei, both lenticular nuclei, the left superior temporoparietal cortex, the left inferior temporal cortex, plus the right superior and right mid-temporal cortex and the left inferior frontal cortex. Reductions of LCBF are patchy, bilateral in every case and, in general, follow the territorial distribution of both middle cerebral arteries. Correlation of LCBF Values With Different Cognitive Functions As Judged By MMQ Subtests Significant correlation coefficients between LCBF values and MMQ subtests categorized according to the different cognitive performances tested are presented in table 4. Left mid-temporal flow reductions correlated with both dyscalculia and memory disturbances. Reduced flow values of both frontal lobes correlated with disorientation to time and place and with attention as tested by repeating and reversing digits and days of the week. The functional contribution of the thalamus to cognitive performance seems to be a broad one as judged by correlation of reduced thalamic LCBF values correlated with MMQ testing. Correlations of thalamic LCBF reductions were significant for impaired calculating ability, problems with attention in repeating and reversing digits, reversing days of the week plus inability to judge and name similarities and opposites. Correlation Between MMQ Scores And Left And Right Temporal Cortex LCBF Values (N=3O) 70 Correlation Of LCBF Reductions With The Severity Of Dementia Table 3 correlates mean LCBF values for different regions with the pooled mean MMQ scores. The severity of dementia correlated directly with reductions of LCBF values for cortical gray matter (r = 0.79, p < 0.01) but not for hemispheric white matter. There were no correlations between MMQ scores and reductions of LCBF values in the caudate and lenticular nuclei. Direct correlations were found between reduced MMQ scores and reduced gray matter flow values measured bilaterally for thalamus and fronto-temporal cortical regions. Correlation coefficients were significant between mean MMQ scores and mean LCBF values for left (r = 0.52, p< 0.05) or right (r = 0 . 7 4 , p < 0 . 0 1 ) frontal cortex, left (r = 0.59, p < 0.05) or right (r = 0.58, p < 0.05) temporal cortex and left (r = 0.62, p < 0.05) or right (r = 0.61, p < 0.05) thalamus (figure 3 and 4). so o •••-L«M r-0.59. P< 0.05 • — n g M r-Q58.p<0.05 40 T w 15 20 MMO Scor. FIGURE 3. Significant correlations were observed between mean MMQ scores and LCBF values for both left and right temporal cortex. STROKE 1006 Correlation Between MMQ Scores And Left And Right Thatamus LCBF Values (N=22) 70 s 60 E o • 50 Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 10 IS L«« r-(l62.p<0.05 Right r - a e i , p < 0 . 0 5 20 MMO Score FIGURE 4. Significant correlations were seen between MMQ scores and left and right thalamus. Discussion It is now generally agreed that CBF is reduced in patients with organic dementia. Regarding average CBF values, O'Brien and Mallet12 and Hachinski, et al14 reported significantly lower CBF values in patients with MID compared to patients with SDAT whereas Ingvar and Gustafson" reported CBF reductions to be about equal in both forms of dementia. Others have obtained results showing that CBF reductions in both MID and SDAT correlate well with the degree of dementia. 1020 Regarding regional flow patterns, Obrist, et al'° and Ingvar, et al18 using the intra-arterial carotid bolus 133Xe method reported local CBF reductions in fronto-temporal regions in SDAT. Yamaguchi, et al20 reported patchy bilateral reductions of CBF in patients with MID utilizing the l33Xe inhalation method. VOL 15, No 6, NOVEMBER-DECEMBER 1984 Computerized reconstruction techniques now make it possible to render images of the brain in three dimensions. Benson, et al39 using PET scanning reported little or no decrease in CMR glucose in patients with MID compared to normal subjects, possibly due to enhanced anaerobic glycolysis. However, studies by Frackowiak, et al40 with oxygen-15 showed that rCBF and rCMR0 2 were markedly depressed in both parietal regions in patients with vascular dementia. Kuhl41 reported scattered asymmetric defects in glucose metabolism in MID. These minor discrepancies reported from PET scanning may be attributed to the relatively poor resolution (circa 1 cu cm) of the method, a disadvantage common to all isotope methods, plus a paucity of cases studied. Current measurements of both LA. and LCBF in patients with MID made by the CTCBF method provide better resolution with more accurate estimates of zones of low flow. Present observations are consistent with earlier reports from this laboratory. Meyer, et al31-42 and Tachibana, et al43 reported multifocal reductions of LCBF in MID with abnormally large coefficients of variation compared to similar measurements in SDAT or in age-matched normals. In considering flow values below which functional disorder may be expected, values below 55 ml/lOOg brain/min for gray matter and below 15 ml for white matter were noted to be associated with cognitive disorders in the present study. These values at which symptoms appear are about 20% below those for agematched normal values.43 Methodological and temporal differences in the measurements of LCBF explain why flow values reported here are higher than critical values associated with neurological deficits in acute experimental studies utilizing implanted hydrogen electrodes.44-45 The present series of patients with MID were allowed to recover from the acute stage of the most recent infarct before LCBF measurements were made, which permits some restoration of CBF to the ischemic areas. Some neurologic and cognitive recovery did occur in some patients, so that some vascular lesions in MID are reversible. Present resolution of TABLE 4 Correlation of LCBF Values with MMQ Subtests which Assess Different Cognitive Performances. Reduced LCBF Values in the Left Mid-temporal Cortex Correlated with Both Dyscalculia and Memory Disturbances. Low Values in Both Frontal Lobes Correlated with Disorientation to Time and Place Cognitive performance tested Orientation Repeating and reversing digits and days of the week Calculation Similarities and opposites Memory Anatomical significant variables (ROI's) Rt. Rt. Lt. Rt. Rt. Lt. Lt. Lt. Rt. Rt. Rt. Rt. Rt. Lt. cingulate cortex frontal cortex frontal cortex cingulate cortex frontal cortex frontal cortex thalamus mid-temporal cortex thalamus frontal cortex sup. temporal cortex thalamus temporal cortex mid-temporal cortex Correlation coefficients 0.76 0.73 0.56 0.83 0.77 0.76 0.62 0.65 0.63 0.62 0.81 0.70 0.56 0.63 Statistical significance p < 0.05 p < 0.05 p < 0.05 p < 0.01 p < 0.01 p < 0.01 p < 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0 05 p < 0.05 p < 0.05 STABLE XENON, CTCBF IN MID/Kitagawa et al Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 CTCBF method utilizing the 8 mm collimator is 80 cu mm, whereas implanted hydrogen electrodes can detect flow changes in regions of 1 cu mm.45 Patients with MID reported here had suffered from their ischemic episodes over a relatively long time interval (over a mean interval of approximately seven years). Chronic lesions may be expected to develop some collateral circulation so that despite persistent functional impairment the flow values are higher than when measured in the acute stages of cerebral infarction.31 In longitudinal follow-up some of our MID cases have shown regions of reversible ischemia with improvement of both MMQ and CBF values which again confirms that some ischemic lesions in MID are reversible. In MID, LCBF values were consistently reduced bilaterally in fronto-temporal cortex and subcortical gray matter and the severity of these reductions correlated with cognitive impairments. Reduced LX values indicate the anatomical site of infarctional or ischemic damage. They are not as useful for making functional correlations as reductions of LCBF values. LCBF values are known to be reduced when there are clinical signs of neurological impairment in the absence of permanent changes of LX values. The functional neurological disorders correlated with reduced LCBF values whether the flow reductions were due to primary ischemia, infarctions or secondary to diaschisis.23-31 The patchy distribution of bilateral ischemic lesions as judged by the map of decreased LA. values is consistent with the topography of hypertensive and lacunar infarcts in the distribution of the middle cerebral artery as reported at autopsy.'"7 Reduction in LA. values in recent ischemia are due to local edema and in old infarcts are due to gliosis which are also associated with low density lesions on plain CT scanning.46 Studies in animal models by CT xenon enhancement have shown that in acute, subacute and chronic infarction, in regions where X values were redued by 10-30 percent, edema followed later by tissue necrosis and gliosis with cystic degeneration was evident. In chronic lesions there were necrotic nerve cells, lipid-laden macrophages, gliosis, and revascularization.46 Reduction of X values in acute ischemia was due to dilution of lipids by water, and in chronic cerebral ischemia was attributed to replacement of normal brain lipids by neutral fats and gliotic tissue that show less solubility for xenon. Xenon is reported to have a Bunsen coefficient of 1.7 in oil at 37° C and 0.085 in water at 37° C with an oil-water ratio of 20. 4? Xenon solubility in water or spinal fluid is low, with near zero values. Short inhalation intervals of'Xe may be insufficient to evaluate true X's especially in diseased areas with low perfusion. However, the computer program described here estimates LA. values extrapolated to saturation at infinity provided LCBF is not below 10 ml/lOOg brain/ min.46 Ladurner, et al8 also reported composite topographic maps of sites of cerebral infarction noted by plain CT scanning in patients with vascular dementia. These maps of low density CT lesions observed in MID are similar to the present topography of reduced LX val- 1007 ues. The main areas affected in both studies were the basal ganglia, thalamus and sylvian cortex bilaterally. The common involvement of the thalamus in MID patients appeared to be an important determinant of dementia in both these studies. The effect of one or more strokes on intelligence varies as a function of location and extent of the infarction. Present results are in agreement with Tomlinson, et al1 who correlated their clinical and neuropathological observations in MID. They concluded that occasionally when the total quantity of destroyed brain tissue was not large, dementia might still be manifested during life if important parts of the brain were destroyed. These authors did not specify which parts of the brain were important in terms of cognitive function other than to mention that they appeared to be in the distribution of the middle cerebral artery. The present study confirms that apart from the anatomical correlates of infarction in patients with MID discussed above, impairment of function in the thalamus bilaterally together with functional disturbance in the cortical distribution of both middle cerebral arteries are important determinants of dementia. While bilateral lesions are necessary, functional impairments within the left temporal cortex appear to play an important role in right handed subjects. Correlation with MMQ subtests suggests that low LCBF values for the left mid-temporal cortex, correlate both with dyscalculia and memory disturbance and that bilateral frontal impairments contribute to disorientation to time and place and to impaired attention. Thalamic ischemia appears to result in overall cognitive blunting. Baer, et al15 also working in this laboratory, utilized a large battery of psychometric tests, including the Wechsler-Adult Intelligence Scale and reported significant correlations between the severity of dementia and mean LCBF values for frontal, temporal and parietal cortex measured by the l33Xe method. This method does not measure flow values in deep brain structures nor permit detailed functional and anatomic cognitive correlations. The CTCBF methods appear to make possible detailed and quantitative correlations between brain topography and function with statistical reproducibility. Within the limitations and constraints imposed by measurements of limited horizontal sections of the brain, present results suggesting that the thalamus is important to cognitive function is not new. Two patterns of dementia have been proposed, a cortical type and a subcortical type.48-49 Cortical dementia is characterized by dysphasia, amnesia and global intellectual deficits but without motor, sensory or affective symptoms which are frequently present in SDAT patients. The concept of subcortical dementia has been based on behavioral syndromes reported in patients with prominent pathological changes in subcortical gray matter. These behavioral syndromes include emotionality and personality change, memory disorders, defective ability to manipulate newly acquired knowledge, and slowing of information processing. Such syndromes were frequently present in our MID patients and presumably STROKE 1008 Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 would be acceptable as examples of subcortical dementia. Emotional incontinence is described as one of the common features included as one criterion in the Hachinski index and aphasic disturbances are also common, although they were excluded in the present study, if severe. Subcortical dementia has been reported in degenerative,50 vascular and neoplas'tic lesions of the thalamus and basal ganglia. Guberman et al51 reported bilateral paramedian thalamic lacunar infarctions associated with subcortical dementia, with characteristic clinical and CT features. Bilateral lesions of the ventrolateral nucleus of the thalamus are associated with modest reductions in scores on the Wechsler Adult Intelligence Scale.49 In general, observations from clinico-pathologic and stereotaxic studies support the hypothesis that thalamo-corticolimbic projection systems participate in cognitive functioning.52'" While the present results support the view that MID constitutes an additional etiology of subcortical "dementia," overt multi-infarct dementia requires additional lesions of frontal and temporal cortex. This is consistent with the view that both cortical and subcortical systems are involved in dementia either directly or by disconnection of projection systems.54 Acknowledgment We wish to acknowledge the assistance of Dr. John Thornby, biostatlstician, VA Medical Center and Baylor College of Medicine in data analysis and Richard E. Dowell, Jr., M.A. for technical assistance. Mrs. Julie E. Link processed the manuscript. References 1. Tomlinson BE, Blessed G, Roth M: Observations of the brains of demented old people. J Neural Sci 11: 205-242, 1970 2. Kay DWK, Foster EM, McKennie AA, Roth M: Mental illness and hospital usage in the elderly. Compr Psychiatry 11: 26-35, 1970 3. BroeGA, AkhtarGR: Neurologic disorders in the elderly at home. J Neurol 39: 362-366, 1976 4. Tower DB: Alzheimer disease — senile dementia and related disorders. Neurological Status. In: Katzman R, Terry RO, Bick KL (eds.) Alzheimer's disease: senile dementia and related disorders (Aging, Vol. 7) Raven Press, New York, 1-4, 1978 5. Hachinski VC, LassenNA, Marshall J: Multi-infarct dementia — a cause of mental deterioration in the elderly. Lancet 2: 207-209, 1974 6. Rosen WG, Terry RD, Fuld PA, Katzman R, Peck A: Pathological verification of ischemic score in differentiation of dementia. Ann Neurol 7: 486-488, 1980 7. Fisher CM: Lacunar strokes and infarcts: A review. Neurology 32: 871-876, 1982 8. Ladumer G, Iliff LD, Sager WD, Lechner H: A clinical approach to vascular (multi-infarct) dementia. In: Meyer JS, Lechner H, Reivich M (eds): Cerebral Vascular Disease: Eleventh International Salzburg Conference, 1982, Amsterdam. Excerpta Medica, 236-243, 1983 9. Lassen NA, Munck O, Tottery ER: Mental function and cerebral oxygen consumption in organic dementia. Arch Neurol Psychiatry 77: 126-133, 1957 10. Obrist WD, Chivian E, Cronqvist S, IngvarDH: Regional cerebral blood flow in senile and presenile dementia. Neurology 20: 315322, 1970 11. Ingvar DH, Gustafson L: Regional cerebral blood flow in organic dementia with early onset. Acta Neurol Scand 46 (Suppl 43): 4273, 1970 12. O'Brien MD, Mallett BL: Cerebral cortex perfusion rates in dementia. J Neurol Neurosurg Psychiatry 33: 497-500, 1970 VOL 15, No 6, NOVEMBER-DECEMBER 1984 13. Simard D, Olesen J, Paulson OB, Lassen NA, Skinhoj E: Regional cerebral blood flow and its regulation in dementia. Brain 94: 273288, 1971 14. Hachinski VC, Iliff LP, Zilhka E, et al: Cerebral blood flow in dementia. Arch Neurol 32: 632-637, 1975 15. Baer PE, Faibish GM, Meyer JS, Mathew NT, Rivera VM: Neuropsychological correlates of hemispheric and regional cerebral blood flow in dementia. In: Meyer JS, Lechner H, Reivich M (eds.): Cerebral Vascular Disease. Seventh International Conference Salzburg. Georg Thieme Publishers, Stuttgart, 100-106, 1976 16. Perez FL, Mattew NT, Stump DA, Meyer JS: Regional cerebral blood flow statistical patterns and psychological performance in multi-infarct dementia and Alzheimer's disease Canad J Neurol Sci 4: 53-62, 1977 17. Meyer JS: Improved method for nomnvasive measurement of regional cerebral blood flow by l33Xe inhalation: Part II. Measurements in health and disease. Stroke 9: 205-210, 1978 18. Ingvar DH, Brun A, Hagberg B, Gustafson L: Regional cerebral blood flow in the dominant hemisphere in confirmed cases of Alzheimer's disease, Pick's disease, and multi-infarct dementia: Relationship to clinical symptomatology and neuropathological findings. In: Katzman R, Terry RD, Bick KL (eds.). Alzheimer's disease: senile dementia and related disorders (Aging, Vol 7), Raven Press, New York, 203-211, 1978 19. Obrist WD: Cerebral circulatory changes in normal aging and dementia. In: Hoffmeister F, Muller C (eds.): Brain Function in Old Age: Bayer Symposium VII, Springer Verlag, 278-287, 1979 20. Yamaguchi F, Meyer JS, Yamamoto M, Sakai F, ShawT Nomnvasive regional cerebral blood measurements in dementia. Arch Neurol 37: 410-418, 1980 21. Butler RW, Dickinson WA, Katholi C, Halsey JH: The comparative effects of organic brain disease on cerebral blood flow and measured intelligence. Ann Neurol 13: 155-159, 1983 22. Lenzi GL, Jones T, Frackowiak RSJ: Positron emission tomography: State of the art in neurology. Prog Nucl Med 7: 118-137, 1981 23. Phelps ME, Mazziotta JC, Huang S: Study of cerebral function with positron computed tomography. J Cereb Blood Flow Metabol 2: 113-162, 1982 24. Meyer JS, HaymanLA, Yamamoto M, Sakai F, NakajimaS: Local cerebral blood flow measured by CT scan after stable xenon inhalation. Am J Neuroradiol 1: 213-225, 1980 25. Meyer JS, Hayman LA, AmanoT, et al' Mapping local blood flow of human brain by CT scanning during stable xenon inhalation. Stroke 12: 426-436, 1981 26. Sakai F, Gotoh F, Ebihara S, Kitagawa Y, Hata T, Takagi Y, Komagamine M: Xenon enhanced CT method for the measurement of local cerebral blood flow in man. J Cereb Blood Flow Metabol 1: 29-30, 1981 27. Ono H, Ono K, Mori K: Mapping of CBF distribution by dynamic Xe-enhanced CT scan method. J Cereb Blood Flow Metabol 1: 5051, 1981 28. Amano T, Meyer JS, Okabe T, Shaw T, Mortel KF: Stable xenon CT cerebral blood flow measurements computed by a single compartment — double integration model in normal aging and dementia. J Comput Assist Tomogr 6: 923-932, 1982 29. Gur D, Wolfson SK, Yonas H, et al: Progress in cerebrovascular disease: local cerebral blood flow by xenon enhanced CT. Stroke 13: 750-758, 1982 30. Segawa H, Wakai S, Tamura A, Yoshimasu N, Nakamura O, Ohta M: Computed tomographic measurement of local cerebral blood flow by xenon enhancement. Stroke 14: 356-362, 1983 31. Meyer JS, Okayasu H, Tachibana H, Okabe T: Stable xenon CT CBF measurements in prevalent cerebrovascular disorders (Stroke). Stroke IS: 80-90, 1984 32. Ackerman RH: Of cerebral blood flow, stroke and SPET. Stroke 15: 1-4, 1984 33. Jacobs JW, Bernhard MR, Delgado A, Strain JJ: Screening for organic mental syndromes in the medically ill. Ann Intern Med 86: 40-46, 1977 34. Folstein MF, Folstein SE, McHugh PR: "Mini-mental state" A practical method for grading the cognitive state of patients for the clinician. J Psychiat Res 12: 189-198, 1975 35. Rosenberg GA, Komfeld M, Stoning J, Bicknell JM: Subcortical STABLE XENON, CTCBF IN MID/Kitagawa et al 36 37. 38. 39. 40. Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 41. 42. 43. arteriosclerotic encephalopathy (Bingwanger): Computerized tomography. Neurology 29: 1102-1106, 1979 Kelcz F, Hilal SK, Hartwell P, Joseph PM: Computed tomographic measurement of the xenon brain-blood partition coefficient and implications for regional cerebral blood flow: A preliminary report. Radiol 127: 385-392, 1978 Rottenberg DA, LuHC, Kearfott KJ: The in vivo autoradiographic measurement of regional cerebral blood flow using stable xenon and computerized tomography: The effects of tissue heterogeneity and computerized tomography noise. J Cereb Blood Flow Metabol 2: 173-178, 1982 Gur D, Shabason L, Wolfson SK, Yonas H, Good WF: Measurement of local cerebral blood flow by xenon-enhanced computerized tomography imaging: A critique of an error assessment. J Cereb Blood How Metabol 3: 133-135, 1983 Benson D, Kuhl D, Phelps M, Cummings JL, Tsia SY: Positron emission computed tomography in the diagnosis of dementia. Ann Neurol 10: 76, 1981 FrackowiakRSJ.PozzilliC, Legg NJ, DuBoulayGH, Marshall J, Lenzi GL, Jones T: Regional cerebral oxygen supply and utilization in dementia. A clinical and physiological study with oxygen15 and positron tomography. Brain 104: 753-778, 1981 Kuhl DE: The effects of aging and stroke on patterns of local cerebral glucose utilization. In: Reivich M (ed.): Cerebrovascular Disease. Thirteenth Princeton Conference. Raven Press, New York, pp 19-27, 1983 Meyer JS, Shaw T, Okayasu H, Tachibana H: Multi-infarct and Alzheimer dementias differentiated from normal aging by xenon contrast CT-CBF measurements. J Cereb Blood Flow Metabol 3 (Suppl 1): S506-S507, 1983 Tachibana H, Meyer JS, Okayasu H, Shaw T, Kandula P, Rogers RL: Xenon contrast CT-CBF scanning of the brain differentiates 44. 45 46 47 48 49 50. 51 52. 53. 54 1009 normal age related changes from multi-infarct dementia (MID) and senile dementia of Alzheimer type (SDAT). J Gerontol 1984, (in press) Jones TH, Morawetz RB, Crowell RM, et al: Thresholds of focal cerebral ischemia in awake monkeys. J Neurosurg 54: 773-782, 1981 Heiss WD: Progress in cerebrovascular disease. Flow thresholds of functional and morphological damage of brain tissue. Stroke 14: 329-331, 1983 Meyer JS, Yamamoto M, Hayman LA, Sakai F, Nakajima S, Armstrong D: Cerebral embolism. Local CBF and edema measured by CT scanning and Xe> inhalation. Neurol Res 2: 101-126, 1980 Cullen SC, Gross EG: The anesthetic properties of xenon in animals and human beings with additional observations on krypton. Science 133: 580-582, 1951 Albert ML. Subcortical dementia. In: Katzraan R, Terry RD, Bick KL (eds.): Alzheimer's disease: senile dementia and related disorders (Aging, Vol 7) Raven Press, New York, 173-180, 1978 Riklan M, Levita E: Subcortical correlates of human behavior. Williams & Wilkins, Baltimore, 1969 Albert ML, Feldman RG, Willis AL: The subcortical dementia of progressive supranuclear palsy. J Neurol Neurosurg Psychiatry 37: 121-130, 1974 Guberman A, Stuss D: The syndrome of bilateral paramedian thalamic infarction. Neurology 33: 540-546, 1983 Fedis P, Van Buren JM: Memory and perceptual deficits during electrical stimulation in the left and right thalamus and parietal subcortex. Brain Lang 2: 78-100, 1975 Brown JW. Language, cognition and the thalamus. Confin Neurol 36: 33-60, 1974 MayeuxR, Stern Y, Rosen J, Benson F. Is "subcortical dementia" a recognizable clinical entity? Ann Neurol 14: 278-283, 1983 Headache in Cerebrovascular Disease RUSSELL K. PORTENOY, M . D . , * t CHRISTOPHER J. ABISSI, M . D . , t RICHARD B. LIPTON, M . D . , * t ALAN R. BERGER, M . D . , * t MARK F. MEBLER, M . D . , t JENNY BAGLIVO, P H . D . , 1 AND SEYMOUR SOLOMON, M . D . * t SUMMARY Two hundred fifteen consecutive patients with cerebrovascular events were evaluated prospectively for the incidence and characteristics of headache. Of 163 patients able to communicate, headache occurred in 29% with bland infarcts, 57% with parenchymal hemorrhage, 36% with transient Lschemic attacks and 17% with lacunar infarcts. Patients with a history of recurrent throbbing headache were significantly more likely to have headache, usually throbbing in quality, during the present illness. Women developed headache significantly more often than men. Headache began prior to the vascular event in 60% of patients and at its onset in 25%. The quality, onset and duration of the headache varied widely among patients. Headache in cerebrovascular disease is common, though neither its occurrence nor characteristics predict lesion type or location. Though the pathogenesis of the headache is unknown, its association with prior throbbing headache suggests that similar factors may operate in both. Stroke Vol 15, No 6, 1984 HEADACHE occurs commonly in cerebrovascular disease. Numerous reports have evaluated headache in patients with ischemic stroke,1"7 cerebral embolism, 2 ' x*• *-'' intracerebral hemorrhage2-3- '• 12~14 and transient ischemic attacks.' 12 ' " • l 6 Headache incidence From the Department of Neurology,* the Headache Unit.t and the Department of Biostatistics,:t Montefiore Medical Center and Albert Einstein College of Medicine. Address correspondence to: Dr. Seymour Solomon, Department of Neurology, Montefiore Medical Center, 111 East 210th Street, Bronx, New York 10467. Received March 15, 1984; revision # 1 accepted May 29, 1984. has varied widely in these series, in part due to differences in patient sampling. Most studies are limited by the problems inherent in the evaluation of transitory symptoms with a retrospective chart review. No prospective study has determined the incidence and characteristics of headache in a single population with such diverse lesions as lacunar and hemispheric infarcts, transient ischemic attacks and parenchymal hemorrhages. In addition, very few surveys have attempted to relate a history of headache to complaints of pain during the acute event. CT-CBF correlations of cognitive deficits in multi-infarct dementia. Y Kitagawa, J S Meyer, H Tachibana, K F Mortel and R L Rogers Stroke. 1984;15:1000-1009 doi: 10.1161/01.STR.15.6.1000 Downloaded from http://stroke.ahajournals.org/ by guest on October 1, 2016 Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 1984 American Heart Association, Inc. All rights reserved. Print ISSN: 0039-2499. Online ISSN: 1524-4628 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://stroke.ahajournals.org/content/15/6/1000 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Stroke is online at: http://stroke.ahajournals.org//subscriptions/