Incidence of Hemorrhagic Stroke in Black Caribbean, Black African, and White Populations
The South London Stroke Register, 1995–2004
Background and Purpose— Data are lacking on the differences in hemorrhagic stroke incidence between black Caribbean (BC), black African (BA), and white ethnic groups. We estimated the incidence for primary intracerebral hemorrhage (PICH) and subarachnoid hemorrhage (SAH) and the associated risk factors for BCs, BAs, and whites.
Methods— First-ever stroke patients were drawn from a prospective community stroke register based in a multiethnic population in South London with 9% BCs, 15% BAs, and 63% whites. Incidence rates were standardized to European and world populations and adjusted for age and sex. Incidence rate ratios (IRRs) relative to whites were calculated by Poisson regression.
Results— Between 1995 and 2004, 566 incident stroke patients were registered: 395 PICHs and 171 SAHs. For PICH, age- and sex-adjusted IRRs were higher in BAs (IRR, 2.80; 95% CI, 2.00 to 3.91) than in BCs (IRR, 1.46; 95% CI, 1.07 to 1.99) and were particularly pronounced for patients age 0 to 64 years: IRR=3.95 (95% CI, 2.65 to 5.87) in BAs and 2.38 (95% CI, 1.50 to 3.80) in BCs. For those <65 years, prestroke hypertension was more prevalent in BAs and BCs (P=0.049). For SAH, the IRR was higher in BCs (IRR; 1.62; 95% CI, 1.05 to 2.48) than in BAs (IRR, 0.80; 95% CI, 0.43 to 1.46).
Conclusions— The higher incidence of PICH observed in BCs and BAs could be explained by prestroke hypertension being more common among young blacks. The different incidences of SAH in BCs and BAs suggest that the baseline risk of stroke for distinct black ethnic groups is not homogeneous.
The incidence of hemorrhagic stroke worldwide varies considerably for both primary intracerebral hemorrhage (PICH) and subarachnoid hemorrhage (SAH).1 Reasons for this finding might be a different risk of hemorrhagic stroke among ethnic groups or a different distribution of main risk factors among the populations.2,3 There is a 2 to 3 times higher incidence in the United States black population compared with whites for PICH and SAH.4–6 Similar ethnic differences in hemorrhagic stroke risk have been observed in previous studies in South London.7 In those studies, the different black groups in the source population were analyzed together. A population study in South London demonstrated that ethnic differences exist in patterns of behavioral risk factors between black Africans (BAs) and black Caribbeans (BCs), with a greater proportion of BAs, for example, having never smoked and a smaller proportion with a heavy alcohol intake compared with BCs.8 Data are lacking on the relation between risk factors and stroke risk for particular ethnic groups.
This study investigated the differences in hemorrhagic stroke risk among whites, BCs, and BAs in the South London Stroke Register (SLSR) and related those differences to prestroke risk factors.
Subjects and Methods
Since January 1995, data have been collected prospectively on first-ever strokes in patients of all age groups in a defined area of South London. At the 2001 census, the total population of the SLSR area was 271 817, with 63% whites; 28% blacks (9% BCs, 15% BAs, and 4% mixed blacks); 5% Asian, Bangladeshi, and Pakistani; and 4% other ethnic subgroups. The number of BAs in the SLSR study area increased from 17 500 to 39 500 between 1991 and 2001, whereas the number of BCs remained stable at ≈26 000 (as per the UK Office for National Statistics). In January 2004, the SLSR area was extended according to the revision of electoral ward boundaries in South London; this increased the population covered by the register by ≈15%.
The detailed methods of notification of patients and data collection for the SLSR have previously been described.9 To maximize case ascertainment, 16 different overlapping sources were used.10 Hemorrhagic strokes were classified as PICH or SAH (including nontraumatic SAH, saccular aneurysms, and arteriovenous malformations) based on results from at least 1 of the following: brain imaging within 30 days of stroke onset (computed tomography or magnetic resonance imaging), cerebrospinal fluid analysis (in all living cases of SAH in whom brain imaging was not diagnostic), or postmortem examination. The diagnosis was confirmed through a review of the clinical information and computed tomography/magnetic resonance imaging scans by a stroke physician (A.G.R.) and a radiologist working independently.
A study clinician performed the initial assessments. Patients were examined as soon as possible after notification to the SLSR. Information collected at the initial assessment included self-definition of ethnic origin (based on the 1991 census question) and socioeconomic status (Registrar General’s codes based on occupation). Ethnic origin was classified into BC, BA, other black, white, and other ethnicity. Socioeconomic status categories were grouped into nonmanual (I, II, and III nonmanual) or manual (III manual, IV, and V) according to the patient’s current (or for the retired and others not currently working, the most recent) employment. An additional category of economically inactive was defined for those with no previous occupation (eg, students) or for whose most recent occupation was unknown. Age was grouped into intervals (0 to 14, 15 to 24, 25 to 34, 35 to 44, 45 to 54, 55 to 64, 65 to 74, 75 to 84, and 85 or older).
Data collected on prestroke risk factors included smoking status, hypertension, diabetes, ischemic heart disease, atrial fibrillation, prescribed antihypertensive drugs, and for diabetics, prescribed hypoglycemic agents/insulin. Patients were considered to be hypertensive if they had records of high blood pressure (>140 mm Hg systolic or >90 mm Hg diastolic) in their general practice or hospital records or if hypertension was diagnosed during the stroke admission. Patients were defined as “treated” for hypertension if, along with a diagnosis of hypertension, there were prestroke records of the prescription of antihypertensive drugs. According to these definitions, patients were divided into 3 groups: untreated hypertension, treated hypertension, and no hypertension.
All cases of hemorrhagic stroke registered between January 1995 and December 2004 were selected for analysis. The SLSR population was estimated from the original study area for each year from 1995 to 2003 by assuming a linear change based on the UK census figures for 1991 (1996 Office for National Statistics adjusted) and 2001. The 2004 population estimate for the enlarged SLSR area was extrapolated in a similar way. The total number of person-years was calculated as the sum of the population figure for each year. This method of person-years estimation was also applied to each component ethnic group.
The BC, BA, and white patients were compared in terms of prestroke risk factors. PICH and SAH cases were analyzed separately. Sex- and ethnic group–specific incidence rates were age adjusted to standard European and world populations.11 Age standardization enables BC, BA, and white populations to be compared according to the same age structure, so that differences in stroke incidence cannot be explained by differences in age distribution. The standard European population is a distribution that approximates the age structure of western European countries. The standard world population is representative of the world as a whole. Similar reasoning applies to sex, as women have greater longevity.
CIs for incidence rate estimates were calculated by the Poisson distribution.12 The incidence rate ratio (IRR) was calculated for BCs and BAs relative to whites by Poisson regression,12 adjusting for age and sex. Interactions between age and sex, age and ethnicity, and sex and ethnicity were included in the models that were considered. For PICH, statistically significant terms of interactions were found for age by ethnicity and age by sex. Thus, results were reported by age strata.
There were 566 patients with a first-ever hemorrhagic stroke (395 PICHs, 171 SAHs). The white, BC, and BA groups are compared in terms of sociodemographic and prestroke risk factors in Tables 1 and 2⇓. For PICH, 37% of whites, 32% of BCs, but only 5% of BAs were smokers (P<0.001). In addition, 60% of whites and 71% of BCs had a history of treated or untreated hypertension compared with 72% of BAs (P=0.053). Hypertension prevalence among ethnic groups mainly differed in patients <65 years old (52% whites, 74% BCs, and 70% BAs; P=0.049). For SAH, 64% of whites were smokers compared with 57% of BAs and 64% of BCs (P=0.858); no other differences among ethnic groups in observed prestroke risk factors approached statistical significance.
Age-specific and age-standardized incidence rates are shown in Table 3⇓. For PICH, crude incidence rates were comparable, but after age adjustment (European or world), BAs had a higher incidence rate than did BCs, who in turn had a higher rate than whites. For SAH, BCs had almost twice the crude incidence rate of whites, with BAs having the lowest rate. After adjustment for age, the much higher rate for BCs remained, but BAs had a rate similar to that of whites.
Table 4 shows the age- and sex-adjusted hemorrhagic stroke IRRs among BCs and BAs compared with whites. For PICH cases, the age- and sex-adjusted IRR relative to whites was higher for BAs (2.80; 95% CI, 2.00 to 3.91) than for BCs (1.46; 95% CI, 1.07 to 1.99). The IRR for PICH in patients age 0 to 64 was 2.38 (95% CI, 1.50 to 3.80) in BCs and 3.95 (95% CI, 2.65 to 5.87) in BAs compared with whites; no statistically significant difference in risk of PICH among ethnic groups was observed in patients age 65 or older. For PICH cases, 2 interactions were found: between sex and age (BC P=0.004, BA P=0.050) and between ethnicity and age (BC P=0.009, BA P=0.007). This latter interaction explains the difference in the results for the 2 age strata.
In SAH, the IRR was 1.62 (95% CI, 1.05 to 2.48) for BCs compared with whites, with BAs having an incidence rate similar to that of whites (IRR, 0.80; 95% CI, 0.43 to 1.46). The risk of SAH was more pronounced in BC patients age 0 to 64 compared with whites (IRR, 1.91; 95% CI, 1.19 to 3.08), but interaction terms were not statistically significant.
In this population-based stroke register, substantial differences in incidence rates and prestroke risk factor levels were found among different ethnic groups. Overall, the risk of PICH was higher in BCs and BAs compared with whites, but this was mainly caused by an increased risk for patients in the age group 0 to 64; in SAH, an excess risk was found only in BCs, again more pronounced in younger patients. The prevalence of prestroke hypertension was substantially increased in BCs and BAs younger than age 65. Smoking patterns seemed not to explain these differences; current smoking had a lower prevalence in BCs and was lower still in BAs.
The overall increased risk of PICH in blacks compared with whites (IRR=1.7) is in accordance with the studies of Broderick et al4 (IRR=1.4) and Flaherty et al5 (IRR=1.8), with the black US group defined by self-identification as black or African American. Our study has found similar associations in BCs and BAs, although the risk of PICH was more pronounced in BAs. The difference in risk for PICH between blacks and whites may be related to different underlying pathologies. PICH can result from a range of different pathologies, including hypertensive damage to arteries, amyloid angiopathy, arterial aneurysm, and arteriovenous malformation. Hypertension can cause PICH either early in the course of the disease, when there has been insufficient time for arteries to hypertrophy, or late, as a consequence of the development of Charcot Bouchard aneurysms in the deep perforating vessels. Lobar hemorrhage in older people is often due to cerebral amyloid angiopathy.2 The BC and BA groups were significantly younger than the whites and might therefore be less likely to have amyloid angiopathy as the cause of stroke.
BC and BA patients showed a higher prevalence of hypertension compared with whites; a more pronounced difference was found in younger patients <65 years of age. Therefore, the observed increased risk of PICH in the black groups might be caused by a higher prevalence of hypertension, as an increased IRR for BCs and BAs compared with whites was found only in patients <65 years. However, we had no information about the prevalence of hypertension in the source population of the SLSR and were therefore unable to adjust for it. Thus, further research on factors explaining the different distribution of hypertension between black and white ethnic groups is needed, including information on the prevalence and management of population blood pressure in stroke-free subjects.
SAH has been shown to be more frequent in BCs than whites but with no significant difference between the BA and white populations. SAH is frequently associated with intracranial aneurysm, and there is some evidence of a genetic link between intracranial aneurysm and SAH.13 SAH has been associated with smoking14,15; however, this does not appear to explain the differences identified in this study, as BC populations in South London showed the lowest prevalence of smoking.
Regarding the observed variations between distinct black groups, a comparison of our findings with the incidence of hemorrhagic stroke and its underlying risk factors in the countries of origin of the different ethnic groups would be useful. However, few studies have investigated the incidence of stroke in low- and middle-income countries.16 The data available for black Africans in the African continent from a single population-based study (Ibadan, Nigeria) give crude rates for PICH and SAH of approximately one fourth of those in the United States and the United Kingdom but with a much younger source population.17 In Barbados, the risk of PICH in BCs is 60% that of BCs living in South London; for SAH, the risk is just 25%.18 There is a lack of data on the prevalence of prestroke risk factors in stroke subjects recruited into good-quality epidemiologic studies from low- and middle-income countries.19 The prevalence of hypertension might vary considerably among indigenous Africans, immigrants to the Caribbean, and black ethnic groups in developed countries. A study by Cruickshank et al20 demonstrated a gradient for age-adjusted hypertension prevalence (defined as 160/95 mm Hg) from 5% in rural Cameroon to 17% in urban Cameroon, 21% in Jamaica, and 29% in Manchester, UK. With hypertension being a major risk factor, particularly for PICH, major differences in incidence should be expected among these settings.
There are strengths and limitations to the study. The strengths of this study are the demonstrated completeness of data collection and case ascertainment in a defined multiethnic population10 and the large numbers of patients, sufficient for subgroup analyses. The weaknesses are that there are limited data on the underlying cause of the hemorrhage. In addition, definitions of prestroke risk factors had to be based on previous medical records, which may have contained inaccuracies or omissions and lacked detail about the effectiveness of treatment of risk factors, particularly hypertension, before stroke.
The higher PICH incidence in BCs and BAs compared with whites might be explained by differences in environmental vascular risk factors, mainly a higher prevalence of prestroke hypertension in young blacks. The differences in the incidence of SAH between BCs and BAs suggest that stroke risk in distinct black communities should not be regarded as homogeneous.
We wish to thank all of the patients and their families and the health care professionals involved. Particular thanks are extended to all of the field workers who have collected data since 1995 for the SLSR. Ilse Burger assisted in obtaining supplementary data from patient notes.
Sources of Funding
This study was supported by the Northern and Yorkshire NHS R&D Programme in Cardiovascular Disease and Stroke, Guy’s and St. Thomas’ Charity, and the Stanley Thomas Johnson Foundation. The authors acknowledge financial support from the Department of Health via the National Institute for Health Research (NIHR) Biomedical Research Centre award to Guy’s & St. Thomas’ NHS Foundation Trust in partnership with King’s College London.
- Received March 7, 2007.
- Revision received April 16, 2007.
- Accepted May 8, 2007.
Feigin VL, Rinkel GJE, Lawes CMM, Algra A, Bennett DA, van Gijn J, Anderson CS. Risk factors for subarachnoid hemorrhage: an updated systematic review of epidemiological studies. Stroke. 2005; 36: 2773–2780.
Flaherty ML, Woo D, Haverbusch M, Sekar P, Khoury J, Sauerbeck L, Moomaw CJ, Schneider A, Kissela B, Kleindorfer D, Broderick JP. Racial variations in location and risk of intracerebral hemorrhage. Stroke. 2005; 36: 934–937.
Quereshi AI, Giles WH, Croft JB. Racial differences in the incidence of intracerebral haemorrhage: effects of blood pressure and education. Neurology. 1999; 52: 1617–1621.
Wolfe CD, Rudd AG, Howard R, Coshall C, Stewart J, Lawrence E, Hajat C, Hillen T. Incidence and case fatality rates of stroke subtypes in a multiethnic population: the South London Stroke Register. J Neurol Neurosurg Psychiatry. 2002; 72: 211–216.
Hajat C, Tilling K, Stewart JA, Lemic-Stojcevic N, Wolfe CDA. Ethnic differences in risk factors for ischemic stroke: a European case-control study. Stroke. 2004; 35: 1562–1567.
Stewart JA, Dundas R, Howard RS, Rudd AG, Wolfe CDA. Ethnic differences in stroke incidence: prospective study using a stroke register. BMJ. 1999; 318: 967–971.
Tilling K, Sterne JA, Wolfe CD. Estimation of the incidence of stroke using a capture-recapture model including covariates. Int J Epidemiol. 2001; 30: 1351–1359.
Armitage P, Berry G, Matthews JNS. Statistical Methods in Medical Research, 4th ed. Oxford, England: Blackwell; 2002: 499–502.
Markus HS, Alberts MJ. Update on genetics of stroke and cerebrovascular disease 2005. Stroke. 2006; 37: 288–290.
Kurth T, Kase CS, Berger K, Schaeffner ES, Buring JE, Gaziano JM. Smoking and the risk of hemorrhagic stroke in men. Stroke. 2003; 34: 1151–1155.
Kurth T, Kase CS, Berger K, Gaziano JM, Cook NR, Buring JE. Smoking and the risk of hemorrhagic stroke in women. Stroke. 2003; 34: 2792–2795.
Osuntokun BO, Bademosi O, Akinkugbe OO, Oyediran ABO, Carlisle R. Incidence of stroke in an African city: results from the stroke registry at Ibadan, Nigeria, 1973–1975. Stroke. 1979; 10: 205–207.
Wolfe CD, Corbin DOC, Smeeton NC, Gay GHE, Rudd AG, Hennis AJ, Wilks RJ, Fraser HS. Estimation of the risk of stroke in black populations in Barbados and South London. Stroke. 2006; 37: 1986–1990.
Cruickshank JK, Mbanya JC, Wilks R, Balkau B, Forrester T, Anderson SG, Mennen L, Forhan A, Riste L, McFarlane-Anderson N. Hypertension in four African-origin populations: current ‘Rule of Halves’, quality of blood pressure control and attributable risk of cardiovascular disease. J Hypertens. 2001; 19: 41–46.