South London Community stroke register (SLSR): Stroke incidence in the first two years of the register (1995-1996)

The aim of this project is to determine whether whether there are ethnic differences in the incidence of first-ever stroke and in the frequencies of different subtypes of stroke. SETTING: An inner-city population of 234,533 residents of Lambeth, Brixton and North Southwark, of which 72 percent are Caucasian, 21 percent are Black (11 percent Afro-Caribbean, 7.5 percent West African, and 2.5 percent Black mixed), and 3 percent are Asian, Bangladeshi and Pakistani. METHODS: Using a multi-source, prospective community stroke register, data on incidence and stroke subtype were collected. Ethnic origin was self-reported by the patient or carer using the standardized OPCS census classification of the group. Stroke subtype was classified using brain radiology (computerized tomography or magnetic resonance imaging) or necropsy data. Strokes without pathological confirmation of stroke subtype using either of these methods were unclassified. RESULTS: A total of 596 strokes were registered in 1995-1996 giving a mean overall annual incidence rate of 1.3 strokes per 1,000 population per year (95 percent C.I. 1.1 -1.4). Incidence of stroke was significanlty higher in all ages in blacks compared with whites (p>0.0001), and the mean age of stroke was significantly lower in blacks compared wih whites (63.5 years, 95 percent C.I. 60.1 - 66.9, and 73.9 years. 95 percent C.I. 74.7 - 72.6, respectively, and p>0.0001). Of the first 500 patients registered, classifications of stroke subtype using information from brain radiology or necroscopy was possible in 441 patients (88.2 percent). There were signfificant ethnic differences in the frequencies of the subtypes of stroke between blacks and whites (p 0.015), with an increased proportion of strokes due to intracranial haemorrhage and lacunar infarction in the blacks (28.4 percent and 29.6 percent) compared with whites (16.4 percent and 24.8 percent), and an increased proportion of non-lacunar infarction in the whites compared with blacks (45.5 percent and 36.4 percent respectively). CONCLUSION: These results suggest important ethnic differences in stroke incidence which probably reflect differences in risk factors for stroke between ethnic groups. These need to be examined further in order to develop effective strategies for stroke prevention in multi-ethnic communities.(AU)

Differences in potentially preventable risk factors for stroke amongst ethnic groups

AIM: To assess the prevalence of risk factors for stroke potentially amenable to health service intervention (hypertension, diabetes, smoking, increased alcohol consumption, physical inactivity, atrial fibrillation, cardiac disease, previous TIA) in black Caribbean, black African and white population of an inner-city health authority. METHODS: A cross-sectional survey was conducted, involving 16 GP Practices serving an area with a high proportion of black residents as identified by the 1991 Census. The FHSA list for these practices was used as the sampling frame. A random sample of 8000 residents (45-74 years old) was selected for a postal survey, which collected data on age, sex, occupation and ethnic group. Responders were stratified by ethnic group. A random sample of 450 subjects in white and black Caribbean group was selected, and together with 193 black African responders invited for screening. RESULTS: There were 725 responders: 303 whites (41.8 percent), 316 black Caribbean (43.6 percent) and 106 black Africans (14.6 percent). Black Caribbeans and Africans were less likely to have a normal blood pressure than whites (OR=0.42, p=0.0001, and OR=0.39, p=0.001 respectively.) Black Caribbeans and were also less likely to have a normal ECG (OR=0.55, p=0.0011, and OR=0.41, p=0.007 respectively.) Left ventricular strain was more common in black Caribbeans (OR=15.81, p=0.008) and Africans (OR=19.97, p=0.007), ischaemic changes were more common in black Caribbeans (OR=2.8, p=0.0001) and myocardial infarction in Africans were found in the prevalence of reported risk factors, such as diabetes (5 percent, 15.2 percent, 10.4 percent respectively, p>0.00001), smoking (31.4 percent, 21 percent, 11.3 percent, p<0.00001, alcohol drinking (80.9 percent, 77.8 percent, 61.3 percent, p<0.00001) and physical activity (75.3 percent, 83.2 percent, 79.3 percent, p=0.048). No difference was found in prevalence of atrial fibrillation and previous TIA. CONCLUSION: Black Caribbeans and black Africans have significantly higher prevalence rates of the important risk factors for stroke. This may partially explain higher mortality rates for stroke in these ethnic groups. Strategies for stroke prevention will be considered in the context of the on-going study on cultural attitudes to risk factor reduction. (AU)