STROKE VOL, No 1, JANUARY-FEBRUARY. Eklof B, Lassen NA, Nilsson L, et al: Regional cerebral blood flow in the rat, measured by the tissue samling technique. Acta hysiol Scand 1: 1-,. Nilsson B: Measurement of overall blood flow and oxygen consumtion in the rat brain. Acta hysiol Scand 2: 12-1,. Gjedde A, Coronna JJ, Hindfelt B, et al: Whole brain blood flow and oxygen metabolism in the rat during nitrous oxide anesthesia. Am J hysiol 22: 3-,. Norberg K, Siesjo BK: Quantitative measurement of blood flow and oxygen consumtion in the rat brain. Acta hysiol Scand 1: 1-1, 1. Rosenblum WI: Effects of reduced hematocrit on erythrocyte velocity and fluorescein transit time in the cerebral microcirculation of the mouse. Circulation Research 2: -3, 1 1. Eklof B, Lassen NA, Nilsson L, et al: Blood flow and metabolic rate for oxygen in the cerebral cortex of the rat. Acta hysiol Scand : -, 3 Role of Factors Associated With Cerebral and Cerebral A Matched air Case-Control Study HASSAN A. H. ABU-ZEID, M.D., D..H., M..H., DR..H.; NUNG W. CHOI, M.D., M..H., H.D.; KULDI K. MAINI, M.SC; ING-HWA HSU, M.S., H.D.; AND NORMA A. NELSON, H.D. Downloaded from htt://ahajournals.org by on November, 201 SUMMARY Comrehensive ascertainment of all ossible new cases of stroke aearing between January 1, 0 and June 30, 1, and admitted to three major hositals in Winnieg, Manitoba, has been achieved by reviewing the Manitoba Health Services Commission claim reorts. The medical records of these cases were reviewed, ertinent data were abstracted, and rigid criteria for diagnosis were followed. Also, data were obtained from death certificates, autosy reorts and long-term hosital records. A total of 0 ascertained cases ( infarction, hemorrhage, and undetermined tye) were matched for age, sex, residence and year of admission with 0 controls from admissions for other than cardiovascular and cerebrovascular disorders. The data were analyzed for elucidating the ossible factors for infarction (INF) and hemorrhage (HGE). FOR CHRONIC INCURABLE DISEASES successful control may be achieved only through rimary revention, and toward this end elucidation of the or causative factors is necessary. Unfortunately, the elucidation of such factors in the case of cerebrovascular disease (stroke) is far from comlete, and reeated descritive and analytic studies are required to uncover those factors and to establish the basis for reventive measures. In revious communications the incidence of cerebrovascular disease by clinical tye and by certain variables such as age, sex, residence, birthlace, occuation and religion was described in a total community including 00,000 residents in Manitoba, Canada. 1-2 Certain etiological hyotheses for the two major tyes of stroke, infarction (INF) and hemorrhage (HGE), were generated. This article resents the results of an analytical case-control study in an attemt to elucidate further factors in stroke and to test the hyotheses generated from the revious studies. An out- From the Section of Eidemiology and Biostatistics, Deartment of Social and reventive Medicine, Faculty of Medicine, University of Manitoba, Room S2A, 0 Bannatyne Avenue, Winnieg, Manitoba, Canada (rerint requests to Dr. Abu-Zeid, Associate rofessor). This study was suorted in art by the Deartment of National Health and Welfare, Canada, Grant Nos. 0--2 and 0--2. The findings suggested that hyertension was the main factor in hemorrhage, whereas in infarction, along with hyertension, other factors such as diabetes, heart enlargement in chest x-ray, ECG abnormalities, and smoking were suggested as factors. There was an association also between infarction, on one hand, and the history of receiving anticoagulants, diuretics, and medications for the heart, and the occurrence of myocardial infarction, on the other hand. These features indicate that infarction and ischemic heart disease have similar factors. Hemoglobin and hematocrit were higher in infarction cases than in their controls only when measured at stroke admission. No difference was revealed when they were measured rior to stroke. Their association with infarction therefore may be secondary to other factors and of no significance for its. standing feature of this design is that it deals only with new cases of stroke and it considers each of the two major entities of stroke (INF and HGE) searately, so that the natural history of each entity might be understood and the relative role of the factors associated with one entity versus the other might be clarified. Methods Virtually all residents in the rovince of Manitoba are registered with the Manitoba Health Services Commission (MHSC) a governmental health care agency which organizes medical and hosital care at no cost for the residents. The atient has the freedom of selecting his own hysician. Since stroke is an acute condition of serious manifestations, and medical care is of no cost to the atient, almost every atient is exected to be brought to the attention of the hysician and subsequently will be hositalized, at least for establishing the diagnosis and for initial care. Thus the MHSC hosital claim reorts rovide a reasonable source for comrehensive ascertainment of all ossible new cases ol stroke aearing in the rovince during a secified eriod From these reorts, all cases diagnosed as cerebrovasculai disease (ISCD rubrics 30-3 Eighth Revision) who wer«
RELATIVE ROLE OF STROKE RISK FACTORS//!bu-Zeid et al. Downloaded from htt://ahajournals.org by on November, 201 admitted to the three major hositals in the greater Winnieg area (about 00,000 residents) Winnieg General Hosital, St. Boniface Hosital, and Victoria General Hosital during the eriod January 1, 0 to June 30, 1, were reviewed. Further detailed data from hosital medical records were obtained by three field research assistants (trained nurses). The data included ersonal characteristics, results of the hysical examination, history of certain diseases, mode of onset, signs, symtoms, and rogress of the case, and laboratory findings. Where available, autosy reorts and data from death certificates and from long-term hosital records also were secured. The data were thoroughly reviewed by one of us (H.A.H.A.) and the criteria of classification and diagnosis of cerebrovascular diseases, established by the ad hoc committee of the Advisory Council for the National Institute of Neurological Diseases and Stroke, 3 were strictly followed to establish a diagnosis for each case. When sufficient evidence was available, the case was assigned to either the INF grou (thrombotic, embolic, or unknown origin) or to the HGE grou (subarachnoid hemorrhage, intracerebral hemorrhage, or hemorrhage of unknown origin). Otherwise, if sufficient evidence were not available to identify the tye of stroke, the case would be designated as stroke of undetermined tye. Only new established cases of acute stroke aearing during the eriod of the study among Manitoba residents and admitted to the three hositals mentioned above were included. Thus 0 cases finally satisfied the criteria for inclusion. These were subclassified by age, sex and residence for selection of controls. Each case was then matched with one control for age (within five years), sex, residence (urban and rural), year of admission, and hosital. The samling frame for the controls included all admissions to the same hositals for conditions other than diseases of the nervous system (ICD 320-3), ischemic heart disease (-1), cerebrovascular disease (30-3), and mental disorders (20-31). For the 0 controls, data similar to those collected for the cases (excet for data related secifically to the condition of stroke such as signs, symtoms and others) were also obtained from the medical records in hositals. The analysis TABLE 1 Differences Between aired Cases and Controls in Some ersonal Characteristics Marital status Religion Birthlace Occuation Relationshi Married, searated, widowed or divorced versus single Jews versus others Mennbnite and Hutterite versus others Catholic versus others rotestant versus others Northern Euroe versus others Eastern Euroe versus others Other Foreign versus others Canada versus others United States versus others Managerial, rofessional, or clerk versus sales service, transortation, rimary or craftsman * control 1 1 2 2 3 2 2 2 1 1 22 was carried out using the matched-air Z-test for testing the difference between cases and controls in case of categorical or discrete variables, and the t-test for testing the difference between the means of the matched airs in case of continuous or numerical variables. Results Of the 0 cases, (.%) were INF, (22.%) HGE, and (.%) strokes of undetermined tye. A total of 33 cases (.%) were from Winnieg General Hosital, 22 cases (3.%) from St. Boniface Hosital and 1 (.%) from Victoria General Hosital. Three hundred forty-six (.1%) were men and 20 (2.%) women. Of the 0 controls, (3%) were admitted for diseases of the digestive and genitourinary systems; (1.%) for endocrinal, nutritional, metabolic, musculoskeletal and skin diseases; 3 (.%) for neolasms; (.%) for diseases of the sense organs; (.%) for diseases of blood and circulatory system other than cerebrovascular and cardiovascular diseases, and 3 (1%) for miscellaneous conditions including accidents, oisoning, violence, comlications of regnancy and uererium, and diseases of the resiratory system. The mean age of cases was. ±1.1 years and for the controls was. ± 1.1 years, indicating a successful match. The results to follow deal with INF and HGE searately. Table 1 shows the comarison between cases and controls with resect to marital status, religion, birthlace and occuation for INF and HGE. airs in which subjects were of unknown status regarding any of these variables were excluded from the analysis. However, among the excluded airs, the attern of differences in these variables was examined between the cases for whom the controls' status was unknown and the controls for whom the cases' status was unknown and found to be similar to that of the airs included in the analysis. As shown in the table there were more of married, searated, divorced, or widowed and less of singles among the cases than among the controls for the INF grou ( = 0.0); no difference was revealed for the HGE 0. 2.0 0. 0.0 Figures between arentheses indicate results based on an exected frequency <. = not significant at a - 0.0. *+ = those who have the characteristics before the word versus, and = those who do not have such characteristics. 2 () 2 1 () 1. 0. (0.) 0. ( ) 1. (0.) 0. () ( ) ()
STROKE VOL, No 1, JANUARY-FEBRUARY TABLE 2 Differences Between aired Cases and Controls in Family History of Cardiovascular Disease and History of Hyertension Family history of stroke Family history of cardiovascular disease Family history of hyertension History of hyertension in atient 1 Control + 2 3.2 2.2 <0.01 control 1 1 0.. Downloaded from htt://ahajournals.org by on November, 201 grou. Aside from the fewer Catholics among HGE cases than among controls ( < 0.02) there were no other observed significant differences in other variables. For comaring the family history or the history of certain diseases between cases and controls, the number of airs of which the case but not the control had such history mentioned in the medical records was comared with the number of airs in which the control but not the case had that history. Table 2 shows these differences in family history of stroke, cardiovascular diseases, hyertension and in the history of hyertension in the atient. In INF but not in HGE the cases had more family history of hyertension than the controls ( = 0.01). In both INF and HGE, however, there was definite history of hyertension in the atients more than in the controls ( ), but no significant difference in the family history of cardiovascular disease or stroke. The medical records were searched for a history or a diagnosis of cardiovascular diseases as well as for a variety of other diseases occurring rior to the stroke for the cases and rior to the condition for which the control was admitted last. Tables 3 and show the frequency of these diseases among the aired cases and controls. It clearly aears that more INF cases than controls had diabetes mellitus ( < 0.0002), myocardial infarction ( ), hyertensive heart disease ( = 0.00), rheumatic heart disease ( ) and embolism ( < 0.0). These differences were not revealed for HGE. On the other hand, in HGE there was a significantly lower frequency of eriheral vascular diseases ( = 0.0), and higher frequency of kidney disease ( = 0.02) and alcoholism ( < 0.03) in the cases than in the controls. In both INF and HGE there was a greater frequency of rior arteriosclerotic heart disease and atrial fibrillation in the cases than in the controls, but no significant difference in the frequency of congestive heart failure and angina ectoris. For other diseases and con- TABLE 3 Condition ditions listed in table no aarent significant differences were revealed excet for the higher frequency of orthoedic disease and other unlisted conditions in the controls than in the cases. The history of receiving secific medications such as antihyertensives, anticoagulants, diuretics, sedatives, and medication for the heart, during admissions rior to stroke and rior to control admission, was comared in both INF and HGE with their controls. Table shows that the antihyertensives and sedatives were more frequently rescribed for the cases than for the controls in both INF and HGE. Whereas only in INF was the revious administration of anticoagulants, medications for the heart and diuretics clearly in excess for the cases than for the controls ( < 0.0002, 0.00,, resectively). Abnormalities in chest x-ray and in the ECG tracings during the first hosital admission rior to the occurrence of stroke or control condition were also checked. These findings are resented in tables and resectively. The mean difference in the eriod between the first hosital admission rior to stroke (or control) and stroke (or control) admission was comared in cases and in controls and no significant difference was revealed. From table it is evident once more that in INF but not in HGE the frequency of detection of heart enlargement in x-ray, ECG abnormalities (articularly ST and T changes) rior to stroke was significantly more in cases than in controls. No difference between cases and controls was detected in the ECG for conduction or rhythm rior to stroke in either INF or HGE. During stroke or control admission, however, both INF and HGE showed a significantly higher frequency of enlarged heart in x-ray, T and ST changes, abnormalities in heart rhythm and generally more ECG abnormalities in the cases than in the controls; the statistical significance was higher in INF than in HGE. The history of smoking and alcohol drinking was not Differences Between aired Cases andcontrolsin Occurrence of Certain Cardiovascular Diseases rior to Stroke or Control MI Angina ectoris Arteriosclerotic heart disease Congestive heart failure Hyertensive heart disease, rior to stroke Rheumatic heart disease, rior eriheral vascular disease, rior Atrial fibrillation, rior Embolism MI = myocardial infarction. Case-h 2 2 1 20 1 0 () 2 1 2 3 3 3 2. 1... 3. (-) <0.00002 0.00 < 0.0002 (<0.0) Controlscase Case-tcontrol () 1 () () () 2 20 0..0 2.0 (.0) (2.0) 0. (-) () 0.02 () () 0.0 (<0.01)
TABLE RELATIVE ROLE OF STROKE RISK FACTORS/A bu-zeid et al. Differences Between aired Cases and Controls in Occurrence of Certain Diseases rior to Stroke or Control Condition Diabetes mellitus Alcoholism Cirrhosis Cancer Gastrointestinal disease Lung disease Kidney disease Orthoedic disease Thyroid disease Hematologic disorder sychiatric disorders Obesity Other diseases 31 2 1 1 2 2 2 2 1 1 2. () 0. 0. 0. 0.3 <0.0002 () <0.03 () 1 1 1 () case- 2 1 30 (-) 0. 0. 2. 0. (2.0) (0.3) 0. 0.2 (<0.03) () 0.02 <0.00 () () Downloaded from htt://ahajournals.org by on November, 201 mentioned in the records in a large roortion of the study grou; however, uon analyzing the airs for which such history was mentioned (table ) it was found that for INF, but not for HGE, there were more smokers among cases than among controls ( ). No significant difference was found in alcohol consumtion for both the INF and the HGE grous. The differences between cases and controls in continuous variables such as blood ressure (B), blood sugar, hemoglobin, hematocrit, body weight and height were tested by alying the t-test for the airs and the results are shown in table. It should be noted that this analysis was carried out only on the airs with available data from the records; any air in which one or both members lacked the data was not included. In general, a reasonably large number of airs rovided information for the above-mentioned variables; however, for cholesterol and triglycerides there were very few data in the records. The analysis of such data was considered unreliable for meaningful results and, therefore, will not be resented here. From table it is evident that systolic B, measured during the first hosital admission rior to stroke or control condition, was significantly higher among INF cases than among controls ( < 0.0), whereas the diastolic B was higher in the case of HGE ( < 0.01). During stroke admission, however, both systolic and diastolic Bs were higher in the cases than in the controls for INF and for HGE ( ). The fasting blood sugar level at stroke or control admission was also higher in cases than in controls for INF ( ) and for HGE ( < 0.0). The data on fasting blood sugar level rior to stroke admission were not available on sufficient cases and controls for analysis. With resect to hemoglobin and hematocrit, there was no significant difference between cases and controls in INF or in HGE when the measurement was made during admission rior to stroke or condition of control. At stroke or control admission, however, hemoglobin and hematocrit levels were significantly higher in cases than in controls for the INF grou only. No difference in body weight and height was observed between cases and controls in both INF and HGE. Discussion It has been shown reviously that the incidence of INF and HGE varied markedly according to age, sex, residence, occuation, religion and birthlace. 1-2 In this reort a fejv findings dealing with religion, occuation and birthlace have not attained statistical significance as that revealed from the incidence data. 2 However, one should kee in mind that the revious reort was based on incidence data relating new cases to the total oulation, whereas in this reort two grous of cases and controls were comared, and the controls were hositalized cases other than stroke or cardiovascular diseases. For certain religions and ethnic grous there is a ossible bias toward admissions to hositals, articularly for the control grou in which there is a good chance that the disease might not have serious manifestations. For this reason we believe that our revious incidence data dealing with these secific variables (religion and birthlace) are more reliable than the case control data. The findings in this study strongly suort the hyothesis that the eidemiological features of INF resemble those of ischemic heart disease. It was shown even that the frequency of occurrence rior to stroke of conditions such as myocardial infarction, hyertensive heart disease, and the frequency of receiving, rior to stroke, antihyertensive drugs, TABLE Antihyertensives Anticoagulants ileart drugs Diuretics sedatives Differences Between aired Cases and Controls in revious Drug Theray in revious Admissions 3 2 31 2 2.2 3. 2.1 1. <0.0002 <0.0002 0.00 0.02 2 1 1 1 1 3. 1. 1. 2. <0.03
1 STROKE VOL, No 1, JANUARY-FEBRUARY TABLE Differences Between aired Cases and Controls in Chest X-ray and ECG Abnormalities, First Hosital Admission Heart enlarged in x-ray ECG, rhythm ECG, conduction ECG, T and ST changes ECG, diagnosis Case-H 2 31 21 2 3 2.1 1. 1. 1. <0.03 0.002 () 0. (0.) () TABLE Differences Between aired Cases and Controls in Chest X-ray and ECG Abnormalities, Stroke and Control Admission Heart enlarged in x-ray ECG, rhythm ECG, conduction ECG, T and ST changes ECG, diagnosis 1 1 12 3 3 2. 2.2 2.3 2. <0.00002 30 2 1 2 3 Control-tease 1 2.3 3.0 1. 1. 3.1 0.00 0.0002 Downloaded from htt://ahajournals.org by on November, 201 anticoagulants, medications for the heart, and diuretics were significantly higher in the INF cases than in the controls. Likewise, during the first hosital admission rior to stroke, the chest x-ray showed a greater frequency of heart enlargement and the ECG showed more occurrences of general abnormalities and of ST and T changes as well as abnormalities in conduction in the INF grou than in the controls. On the other hand, in HGE, excet for the significantly higher frequency of receiving antihyertensives rior to stroke by the cases than by the controls, none of the abovementioned abnormalities in chest x-ray and ECG, or of the frequency of occurrence of diseases or receiving medications, were significant. This indicates that hyertension is a common factor in INF and in HGE; but other features of ischemic heart disease are commonly shared with INF only. The occurrence of rheumatic heart disease, atrial fibrillation, and embolism more in INF cases than in controls exlains the factors related to the embolic comonent of the INF grou. It is interesting to note that during stroke admission manifestations of heart disease such as heart enlargement in x-ray, ECG abnormalities in rhythm, T, ST, and general ECG abnormalities were also significantly more evident in HGE cases than in controls. Such abnormalities did not exist during hositalization rior to stroke. This indicates that in HGE, heart abnormalities occur very closely or concomitant with the occurrence of stroke, whereas in INF, such abnormalities aeared rior to the stroke indicating the robability of their being factors in INF or that both share common factors. This study also confirms the very imortant role of high blood ressure in the genesis of both INF and HGE. The observed significantly high systolic B in INF and high diastolic B in HGE rior to stroke may suggest a secific role of each in the genesis of these conditions. However, the differences in diastolic B for INF and in systolic B for HGE were of borderline significance (0.0 < < 0.); also, both systolic and diastolic Bs were equally significantly higher in the cases than in the controls in either INF or HGE when measured at stroke admission. Hyertension has been labeled universally as the most common and otent recursor of INF. " It has been shown that its contribution is direct and indeendent of other factors. Hyertension is known to be a factor in the genesis of cerebral atherosclerosis, " and this may exlain its role in INF. With regard to HGE, hyertension is singled out as the main associated factor, and some studies including ours showed that this association was stronger than in the case of INF. 2 - Hyertension may also lead to heart enlargement and to ECG abnormalities. The fact that such abnormalities aeared sometime before the occurrence of INF but aeared close to the occurrence of HGE in this study ma> indicate that the level of blood ressure might have beer elevated for a long eriod in INF, whereas the elevation was for a shorter eriod in HGE. The clear excess of diabetes mellitus in INF but not ir HGE suorts its role in the develoment of INF. Such i role also has been shown from the Framingham data ir which an increased of INF was observed in ersons witl TABLE Smoking Alcohol drinking Differences Between aired Cases and Controls in Smoking and Alcohol Consumtion case 3 1 2. 1. Case-Icon trol () ()
RELATIVE ROLE OF STROKE RISK FACTORS/Abu-Zeid et al. Ill TABLE Differences Between aired Cases and Controls in Blood ressure, Fasting Blood Sugar, Hemoglobin and Hematocrit Levels Systolic B 1 Diastolic B 1 Systolic B 2 Diastolic B 2 Fasting blood sugar 2 Hemoglobin 1 Hemoglobin 2 Hematocrit 1 Hematocrit 2 Height (cm) Weight (kg) No. airs 03 3 2 1 30 11 Mean difference*.2 3.2 1..00 20...3-2.1-0. - SE 3.0 2.01 2.1 2. 3.0 2.21 2.1 0. <0.0 No. airs 3 3 3 3 22 Mean difference* SE..1 3. 1.3.3-3.. -.31 1. 3.2 -..3 3.0.01 2.2 2.2 2.0.3 0.1 2.1. Mean difference comutation based on the measurement of the case minus the measurement of the control for the airs. 1 = First hosital admission,! = stroke admission. <0.01 <0.0 Downloaded from htt://ahajournals.org by on November, 201 even modest evidence of imaired glucose tolerance. The observed unusually high level of fasting blood sugar for INF in this study and the imaired glucose tolerance among stroke candidates reorted in other studies ' ~ 1 rovide further evidence for the same thesis. That diabetes or imaired glucose tolerance increases the for INF may result from its robable contribution to cerebral atherosclerosis, for autosy studies have shown that diabetes is associated with an increased frequency of cerebral atherosclerosis. 1 ' " It has been shown in our data also that the fasting blood sugar level in HGE was high. This, however, may be a transient hyerglycemia which may result from disturbed sugar metabolism due to cerebral injury. 3 Further suort for this is rovided from the observation that diabetes as such was not associated with HGE as it was with INF. From this study it seems also that the smoking habit is associated with INF but not with HGE. Along with the observed high level of blood ressure and more occurrence of diabetes in the INF grou, smoking may initiate or enhance the atherogenic rocess in cerebral vessels. No evidence has been found in this series to indicate any association between body weight and stroke. Results similar to ours in this regard were reorted from rosective data. ' The high level of hemoglobin and hematocrit observed in INF but not in HGE is of interest and warrants further consideration. The athohysiologic mechanism of these factors in the genesis of INF may involve high blood viscosity and trauma to the intimal lining of blood vessels, ossibly due to larger, heavier, or increased number of red blood cells than usual. These factors also may lead to thrombosis. The Framingham study has initially shown that the of develoing INF among those at the high scale of the normal range of hemoglobin was significantly higher than among those at the lower end of the scale. However, when the data were adjusted for hyertension a factor related to both hemoglobin and stroke the residual was small and not significant. Thus, the association between INF and hemoglobin or hematocrit may be a secondary one and of no causal significance. The same relationshi was also revealed from cross-sectional data between hemoglobin and ischemic heart disease factors; 20 also, this association was later shown from rosective data to be of no significance in the genesis of this disease. 21 It was reorted reviously that the incidence of INF was significantly higher in men than in women of urban areas, and, among men, those of urban areas had a higher incidence than those of rural areas. 1-2 For HGE, the incidence was higher in women than in men of urban areas, and, among women, the incidence was higher in those of urban than of rural areas. Based on this attern it was suggested that the revalence of ossible factors such as hyertension, high serum liids and smoking may vary according to sex and residence, and may exlain this attern of incidence for INF and for HGE. The analysis of the available data for the control grou used in this study artly suorts this hyothesis. The mean systolic B was significantly higher in women than in men (1. mm Hg versus.2, < 0.02); this was articularly evident among residents of urban areas (1. mm Hg for women versus.3 for men, < 0.01). The diastolic B was also higher in women of urban areas than in those of rural areas (.2 mm Hg versus, ). These atterns in blood ressure concur with the incidence atterns of HGE by sex and residence. The data from medical records on serum liids and smoking were very few for the control grou and, therefore, were not resented. However, the results from this case-control study along with our revious descritive incidence results indicate that high blood ressure is so far the main factor in HGE, whereas in INF high blood ressure and other factors associated with the genesis of ischemic heart disease all aear to be imortant. Acknowledgments We wish to thank Dr.. D. Henteleff, Mr. F. Toll, and the other staff of the Manitoba Health Services Commission, who generously cooerated in making the imlementation of this study ossible; the medical records deartments of various hositals in Manitoba for their ermission to review the records; Mrs. E. Austen, Miss A. Delaney and Mrs. W. Michaluk for the field work; and Mr. Y. S. Youn for heling in the analysis and tabulation. References 1. Abu-Zeid HAH, Choi NW, Nelson NA: Eidemiologic features of cerebrovascular disease in Manitoba: Incidence by age, sex and residence, with etiologic imlications. Canad Med Assoc J 3: 3-3, 2. Abu-Zeid HAH, Choi NW, Maini KK, et al: Incidence and eidemiologic features of cerebrovascular diseases (stroke) in Manitoba, Canada. Study of oulation 20- years of age. rev Med : -, 3. Millikan CH, Adams RD, Fang H, et al: A classification and outline of
2 STROKE VOL, No 1, JANUARY-FEBRUARY cerebrovascular diseases. Neurology : 3-3,. Armitage : Statistical Methods in Medical Research. New York, John Wiley and Sons, and,. Kannel WB, Wolf A, Verter J, et al: Eidemiologic assessment of the role of blood ressure in stroke. The Framingham study. JAMA 21: 301-3, 0. Gertler MM, Rosenberger JL, Leetma AH: Identification of individuals with covert ischemic thrombotic cerebrovascular disease: A discriminant function analysis. Stroke 3: -1, 2. Chaman JM, Reeder LG, Raymond B, et al: Eidemiology of vascular lesions affecting the central nervous system: The occurrence of strokes in a samle oulation under observation for cardiovascular diseases. Am J ublic Health : 1-201,. William AO, Loewenson RB, Liert DM, et al: Cerebral atherosclerosis and its relation to selected diseases in Nigerians: A athological study. Stroke : 3-01,. Baker AB, Resch JA, Loewenson RB: Hyertension and cerebral atherosclerosis. Circulation 3: 01-,. Solberg LA, McGarry A: Cerebral atherosclerosis in Negroes and Caucasians. Atherosclerosis 1: 11-1, 2. Kannel WB: Current status of the eidemiology of brain infarction associated with occlusive arterial disease. Stroke 2: 2-31, 1. Gertler MM, Leetma HE, Koutrouly RJ, et al: The assessment of insulin, glucose and liids in ischemic thrombotic cerebrovascular disease. Stroke : -,. Gertler MM, Rush HA, Whiter HH, et al: Ischemic cerebrovascular disease: The assessment of factors. Geriatrics : -11, 1. Albanese AA, Lorenze EJ, Orto LA: Effect of strokes on carbohydrate tolerance. Geriatrics : 12-10, 1. Jacobson T: Glucose tolerance and serum liid levels in atients with cerebrovascular disease. Acta Med Scand : 3-23, 1. Bell ET: A ostmortem study of vascular disease in diabetes. Arch ath 3: -, 2 1. Grunnet ML: Cerebrovascular disease, diabetes and cerebral atherosclerosis. Neurology : -1, 3 1. Kannel WB, Dawber TR, Cohen ME.etal: Vascular disease of the brain: Eidemiologic assessment. The Framingham study. Am J ublic Health : -,. Kannel WB, Gordon T, Wolf A, et al: Hemoglobin and the of cerebral infarction: The Framingham study. Stroke 3: 0-20, 2 20. Abu-Zeid HAH, Chaman JM: The relationshi between hemoglobin level and some factors in ischemic heart disease. The Los Angeles heart study. Circulation (Sul IV):, 3 21. Abu-Zeid HAH, Chaman JM: The relation between hemoglobin level and the for ischemic heart disease: A rosective study. J Chron Dis 2: 3-03, Reserine and Cerebral Vasosasm CLARK WATTS, M.D. Downloaded from htt://ahajournals.org by on November, 201 SUMMARY Cerebral vasosasm was roduced in the dog basilar artery by toically alied five-day-old clotted autologous blood, but not by freshly drawn blood. The sasm was reversed by methysergide, an antiserotonin agent. However, vasosasm was not roduced by five-day-old clotted autologous blood from Introduction CEREBRAL VASOSASM often comlicates the management of atients with intracranial aneurysms and is usually associated with blood in the subarachnoid sace. 1 It has been concluded that the redominant vasoconstrictor elements of blood are contained in the latelet fraction. 2 The storage in latelets of certain vasoactive amines liberated during normal clotting is influenced by reserine, 3 " a hoshodiesterase inhibitor. The resent study was undertaken to ascertain the vasoactive effect uon the basilar artery of autologous blood from dogs retreated with reserine. Methods The basilar artery was exosed in mongrel dogs using a transclival aroach under intravenous entobarbital anesthesia with endotracheal intubation and sontaneous, unassisted resiration. A Zeiss oerating microscoe was used to oen the dura and remove the arachnoid membrane. Observations of the ventral brain stem and the basilar artery and its branches were made continuously through the oerating microscoe. The changing diameter of the basilar artery was measured through the microscoe by a micrometer calibrated to the nearest 0.1 mm. There were three exerimental grous. Grou A con- rofessor of Surgery and Chief, Section of Neurosurgery, University of Missouri Medical Center, Columbia, Missouri 201. dogs retreated with reserine. This suggests latelet serotonin or a similar, unidentified substance as the vasosastic element in dog blood resonsible for exerimental vasosasm from toically alied whole blood. Other exerimental data suort these findings. tained three dogs. The basilar artery of each dog was bathed in fresh autologous blood for 30 minutes and then washed briefly with normal saline at 3. C. The diameter of the artery was measured immediately before and after treatment. Grou B contained six dogs. They were treated as were the dogs in Grou A. The blood used, however, had been drawn from the secific animal tested five days rior to the exeriment, allowed to clot, and then stored at 3. C until used in the exeriment. After the blood was irrigated from around the vessel and the osttreatment diameter was measured, the vessels of three dogs were bathed in a solution of methysergide (1 mg/cc) for 1 minutes. The diameter was again measured. There were four dogs in Grou C. Each dog received reserine, 0.1 mg er kilogram intramuscularly, daily for four days. Then blood was drawn from each animal, allowed to clot, and stored at 3. C for five days. At the end of that eriod, the basilar artery of each dog was exosed as noted above and then treated with autologous blood, as in Grou A. Results The results are summarized in table 1. In Grou A, no vasosasm was roduced in any animal by the treatment of the basilar artery with fresh autologous blood. Vasosasm was seen in the basilar artery of each animal in Grou B. These vessels were treated with five-day-old