Impact of Race and Ethnicity on Ischemic Stroke
The University of California at San Diego Stroke Data Bank
Background and Purpose As the US minority population continues to grow, increasing numbers of nonwhite citizens are at risk for stroke. A better understanding of how ischemic stroke differs in the minority populations may lead to more effective clinical management.
Methods We prospectively evaluated 542 consecutive patients (416 whites, 71 Mexican Americans, 55 blacks) presenting to the University of California at San Diego Medical Center or the San Diego Veterans Affairs Hospital with presumed acute ischemic stroke or transient ischemic attack.
Results Whites had a higher proportion of transient ischemic attacks (32% versus 18% and 17% for blacks and Mexican Americans, respectively) and had the lowest prevalence of diabetes mellitus (17% versus 29% and 40% for blacks and Mexican Americans, respectively). Mexican Americans had higher initial serum glucose levels (178 versus 133 and 131 mg/dL for whites and blacks, respectively). Blacks were youngest (average age, 56 years). There were no differences among the groups in the prevalence of prior stroke, hypertension, myocardial infarction, or smoking; initial systolic blood pressure, serum cholesterol levels, and functional deficit also were similar. Although it did not reach statistical significance, there was a trend toward relatively late presentation in the black stroke subpopulation: only 53% of blacks (compared with 73% of both Mexican Americans and whites) reached medical attention within 24 hours of stroke onset. All groups had similar diagnostic evaluations and functional outcome at 1 week. With the exception of a higher frequency of stroke of unknown cause in Hispanics, the distributions of stroke etiologies did not differ significantly among the groups.
Conclusions These data suggest that there are significant clinical differences in populations with ischemic stroke and transient ischemic attack that are related to race and ethnic origin, but in our population these differences did not include the extent of diagnostic evaluation undertaken or stroke severity.
The US minority population is growing and may constitute 25% of the country’s total population by the year 2000. A better understanding of the clinical features of cerebrovascular disease in this population is required. Most prior studies of racial/ethnic differences in stroke have involved comparisons of the prevalence of atherosclerotic risk factors in whites versus nonwhites,1 and few data exist regarding specific minority subpopulations; even less is known regarding the impact of racial/ethnic origin on the clinical presentation and evaluation of the stroke patient. This is especially true of the Hispanic subpopulation, which accounts for 9% of the US population and 20% of the San Diego County population.2 We undertook this study to determine whether clinical differences related to race/ethnicity might exist in our ischemic stroke and transient ischemic attack (TIA) population.
Subjects and Methods
Over a 5-year period beginning January 1, 1984, and concluding December 31, 1988, we prospectively evaluated a consecutive series of patients presenting to the University of California at San Diego (UCSD) Medical Center or the San Diego Veterans Affairs Hospital with presumed acute ischemic stroke or TIA; data relevant to the patients were entered into a computerized registry. Those data included details of the patient’s past medical history, acute stroke/TIA presentation, physical examination, diagnostic studies, treatment, and clinical course. When possible, we obtained historical data through direct patient interview; alternatively, we questioned the family or performed a chart review. The criteria defining a history of hypertension were (1) a self-reported history, (2) current use of antihypertensive therapy, or (3) a chart review yielding three or more readings of systolic pressure ≥165 mm Hg or diastolic pressure ≥95 mm Hg. A history of diabetes was defined either by self-report or by the current use of a hypoglycemic agent.
After completion of the diagnostic evaluation, we assigned each stroke a specific etiology according to predetermined diagnostic criteria that have been summarized in detail elsewhere.3 Briefly, a diagnosis of large-vessel atheroembolic stroke required evidence by arterial imaging of stenosis or occlusion in an appropriate vessel, and diagnosis of lacunar stroke required that imaging studies be performed and demonstrate no such large-vessel disease; diagnosis of cardioembolic stroke required the presence of a cardiac condition with recognized high embolic potential and the absence of intrinsic cervical or cranial large-vessel disease, as evidenced by arterial imaging.
We reexamined all patients at 1 week or at the time of hospital discharge, according to which came earlier. We used a compressed Rankin scale to assess each surviving patient’s functional status at initial presentation and at follow-up, with ratings of normal (ie, no new neurological deficit), mildly impaired (ie, new deficit but can return to work or usual baseline level of daily functioning), moderately impaired (ie, unable to return to work or usual baseline level of functioning but does not require chronic institutionalization or assistance for dressing, cooking, eating, or bowel/bladder functions), and severely impaired (ie, requires chronic institutionalization and/or chronic assistance for basic personal needs).
We categorized our patients based on the subjective judgment of one of the investigators as white, black, Mexican American, or other. Individuals with only one parent of nonwhite origin were given the racial/ethnic origin of that parent. Variables to be examined in these racial/ethnic subgroups included stroke etiology; prevalences of hypertension, diabetes, prior stroke, prior myocardial infarction, and smoking; frequencies with which certain diagnostic studies (echocardiogram, carotid duplex, brain computed tomogram, brain magnetic resonance imaging, cerebral arteriography) were performed; time interval from stroke or TIA onset to initial medical presentation; and functional status at presentation and at 1 week.
In comparing the patient groups for differences, we defined statistical significance as P<.05 using ANOVA with the Newman-Keuls test for continuous data. We used the χ2 test with Bonferroni correction for multiple comparisons for proportions.
We evaluated 575 patients with stroke, TIA and stroke, or TIA alone. There were 416 whites, 71 Mexican Americans, 55 blacks, and 33 of other racial/ethnic origin. Because in this last category no single group was sufficiently represented to allow meaningful comparisons, we restricted our analyses to the first three categories. Of these 542 patients, 386 had stroke, with or without temporally associated TIA, and 156 had TIAs only. The proportion of whites with TIA only (134/416=32%) was higher than that of blacks (10/55=18%; P=.06) or Mexican Americans (12/71=17%; P=.02). Blacks were significantly younger than whites, with Mexican Americans occupying an intermediate position, and there was a significantly higher proportion of females in the black group than in the white (Table 1⇓). The only significant difference in the distributions of stroke risk factors was that, relative to whites, more Mexican Americans were diabetic. There were no significant differences in the prevalence of hypertension, history of previous stroke or myocardial infarction, or active tobacco use among the three groups; the prevalence of hypertension in patients aged 55 years or younger did not significantly differ among the groups (whites 37%, blacks 48%, Mexican Americans 48%).
Mexican Americans had a higher initial mean serum glucose concentration than either whites or blacks (Table 2⇓). There was no significant difference in initial mean systolic blood pressure or cholesterol concentration. Fewer blacks or Mexican Americans than whites presented within 6 hours of stroke onset, but these differences did not reach statistical significance. The groups received similar diagnostic evaluations. Among the stroke patients, no differences were found in functional status at presentation or at 1 week.
Regarding distribution of stroke etiologies among the three groups, there was only one significant difference: compared with whites, Mexican Americans had more strokes of unknown etiology (Table 3⇓).
Prior studies involving stroke as it relates to racial/ethnic origin largely have been restricted to comparisons of black and white populations.1 2 3 4 Our nation’s Hispanic subpopulation is growing rapidly, and information regarding cerebrovascular disease in this group is badly needed. Studies of Hispanics may have limited generalizability: the Hispanic subpopulation is heterogeneous, and differences between various subgroups (Mexican versus Cuban versus Puerto Rican, etc) may exist. What investigators intend when speaking of a “Hispanic” study should be strictly defined; in this study, our Hispanic population was entirely Mexican American.
Racial/ethnic origin may influence the clinical characteristics of ischemic stroke. We found a higher percentage of white patients presenting with TIA only. This may indicate a racial/ethnic difference in the pathophysiology of cerebrovascular disease itself, the threshold for seeking medical attention, or the availability of medical care. That our black stroke/TIA patients were younger than either whites or Mexican Americans may reflect Caplan’s5 finding of a higher stroke risk in younger black patients relative to whites. Alternatively, this finding may be simply an artifact related to the two institutional sources of our patient population: a Veterans Affairs Hospital (yielding patients relatively more elderly and more often white) versus a university medical center (providing the bulk of our younger, and more often black and Mexican American, patients). Black stroke patients have been reported to have higher frequencies of hypertension, diabetes, heart disease, and smoking than white patients,4 6 findings we could not duplicate.
Little is known concerning the prevalence of stroke risk factors in Mexican Americans.7 Studies of samples from the general population have shown a higher prevalence of diabetes in Mexican Americans than in whites,8 9 but only two studies have examined the risk-factor profiles of Hispanic stroke patients.10 11 Sacco et al10 found increased prevalences of hypertension and diabetes in their Northern Manhattan Hispanic stroke patients relative to blacks and whites. Frey et al11 similarly noted a higher frequency of hypertension and diabetes in their Hispanic stroke population relative to whites. As we did not control for the rate of diabetes in our general population, we cannot draw any definite conclusions regarding what contribution diabetes may make toward stroke risk in Mexican Americans. Even so, extrapolating from the other series that have been reported, one can hypothesize that the relative excess of diabetes and elevation of acute serum glucose levels found in our Mexican American stroke patients most likely parallel the higher prevalence of diabetes found in the Mexican American population at large.8 9 We found no other statistically significant differences in major stroke risk factors, but this may have been due to the relatively small numbers of patients in the minority groups.
With the advent of effective therapies for stroke prevention after TIA and, potentially, for acute stroke itself, the time interval from event onset to medical presentation has become increasingly important. Although it did not reach statistical significance, our data indicate a trend toward relatively late presentation in the black stroke subpopulation; only 53% of blacks (compared with 73% of both Mexican Americans and whites) reached medical attention within 24 hours of stroke onset. The lack of significance possibly represents a type II error caused by the relatively small number of black stroke patients in our sample. Whether public education efforts directed at specific subpopulations may reduce this time interval is unknown, but some evidence suggests that such is the case, at least for the general population. Alberts et al12 13 reported an increase (from 36% to 86%) in the proportion of patients presenting within 24 hours subsequent to the implementation of a local public and professional education program.
Although we found only limited differences in the etiologies of ischemic stroke among our subpopulations, others have reported a greater proportion of lacunar strokes in blacks and more cardioembolic strokes in whites.6 14 There are few data regarding the distribution of ischemic stroke etiologies in Mexican Americans. The higher proportion of stroke of unknown cause in our Mexican American patients is not easily explained. The degree of diagnostic evaluation attempted was not obviously less in this group, but bias induced by the diagnostic criteria we used may have served to push a number of potentially identifiable etiologies into the “unknown cause” category. Alternatively, a higher proportion of ischemic strokes in Mexican Americans may be due to another process, yet to be identified, such as hypercoagulability. Other investigators have demonstrated differences in blood coagulability related to race.15 Further studies investigating stroke of unknown etiology in general and in relation to race/ethnicity are needed.
There is little known regarding the influence of racial/ethnic origin on ischemic stroke severity. We found no differences in ischemic stroke mortality or morbidity at presentation or at 1 week among the three groups. Our findings are similar to those from the Lehigh Valley Study, which measured in-hospital mortality for blacks and whites, and to those from a study of hospitalized whites, blacks, and Hispanics in Northern Manhattan.6 10 16 Also consistent with our results, Frey et al11 reported no difference in in-hospital death rates in a population of whites, Hispanics, and Native Americans with acute stroke.
Even with the relatively small numbers of minority patients involved, our results suggest significant clinical differences in ischemic stroke/TIA populations that are related to racial/ethnic origin. We found additional differences that failed to reach statistical significance, and this quite likely reflected our small sample size. A better understanding of these differences may lead eventually to more effective and population-specific therapies for both the prevention and treatment of stroke.
This research was supported in part by a grant from the American Heart Association, San Diego Chapter and by a Department of Veterans Affairs Fellowship in Neurosciences and Traumatic Brain Injuries (Dr Zweifler).
Data from this project were presented at the World Congress of Neurology Conference (Vancouver, Canada, September 7, 1993) and published in abstract form in the Canadian Journal of Neurological Sciences.
- Received August 12, 1994.
- Revision received November 16, 1994.
- Accepted November 17, 1994.
- Copyright © 1995 by American Heart Association
Caplan LR, Gorelick PB, Hier DB. Race, sex and occlusive cerebrovascular disease: a review. Stroke. 1986;17:648-655.
United States Bureau of the Census. 1990 Census of Population and Housing—Summary: United States. Washington, DC: US Bureau of the Census; 1992:59-82.
Caplan LR. Strokes in African-Americans. Circulation. 1991;83: 1469-1471.
Friday G, Lai SM, Alter M, Sobel E, LaRue L, Gil-Peralta A, McCoy RL, Levitt LP, Isack T. Stroke in the Lehigh Valley: racial/ethnic differences. Neurology. 1989;39:1165-1168.
Yatsu FM. Strokes in Asians and Pacific-Islanders, Hispanics, and Native Americans. Circulation. 1991;83:1471-1472.
Hanis CL, Ferrell RE, Barton SA, Aguilar L, Garza-Ibarra A, Tulloch BR, Garcia CA, Schull WJ. Diabetes among Mexican Americans in Starr County, Texas. Am J Epidemiol. 1983;118:659-672.
Stern MP, Gaskill SP, Allen CR, Garza V, Gonzales JL, Waldrop RH. Cardiovascular risk factors in Mexican Americans in Laredo, Texas. Am J Epidemiol. 1981;113:546-555.
Sacco RL, Hauser WA, Mohr JP. Hospitalized stroke in blacks and Hispanics in Northern Manhattan. Stroke. 1991;22:1491-1496.
Frey J, Jahnke H, Giles S, Boston M, Feinberg W. Epidemiological features of stroke in white, Hispanic and Native American populations of Arizona. Neurology. 1993;43:A395. Abstract.
Alberts MJ, Perry A, Dawson DV, Bertels C. Effects of public and professional education on reducing the delay in presentation and referral of stroke patients. Stroke. 1991;23:352-356.
Sacco RL, Hauser WA, Mohr JP, Foulkes MA. One-year outcome after cerebral infarction in whites, blacks, and Hispanics. Stroke. 1991;22:305-311.