Stroke: ethnic differences do exist

Stroke is a major preventable cardiovascular outcome that is responsible for a high proportion of worldwide morbidity and mortality. Accounting for nearly 10% of all deaths, stroke mortality rates are second only to coronary heart disease (13%) and cancer (12%) [1]. Traditional research into etiology, prevalence and mortality have been the key areas of stroke research but, more recently, there has been a global interest in treatment and prevention strategies, outcome measures, and secular trends in stroke. We know that there is a geographical and regional variation in stroke incidence and prevalence, but clear evidence to explain the basis of this variation is lacking. Potential explanations include differences in risk factor prevalence, socioeconomic status (accounting for disparate treatment), and environmental and genetic factors. Although no study has systematically examined differences in the effect of risk factors in different race–ethnic groups, the literature has consistently shown ethnic variations in the prevalence of cardiovascular risk factors [1].

The changing demographic profile of the developed and developing countries with increasing numbers of elderly and minority groups has profound implications on our current approach to stroke risk stratification. Despite the mounting evidence that stroke risk factors may differ among racial and ethnic groups, public health programs and national guidelines for stroke prevention continue to focus on the identification of risk factors in the general population without addressing the specific needs of ethnic groups. Definitive differences among ethnic groups are best measured through a comparison of individuals from the same population or community, thus minimizing bias that may arise from comparisons of groups living in different physical, social and healthcare environments. For example, although black/African–American subjects in the USA have a higher incidence of stroke, more severe strokes and higher stroke mortality than whites, it is also well known that blacks have a lower socioeconomic status and income, both of which are independent risk factors for stroke [2]. In addition, there is a higher prevalence of recurrent stroke among Hispanic than white patients [3]; however, the studies are hampered by the heterogeneity among Hispanics who originate from Mexico, South America and the Caribbean basin, each bringing along with them a variety of race and ethnic-specific risk factors that have not been previously taken into account.

Similarly, in Britain, there are marked differences in the distribution of cardiovascular risk factors, and their detection and management varies among different ethnic groups with a higher incidence of diabetes and hypertension among South Asians, Caribbeans and West Africans, who may be at a lower income and socioeconomic status [4]. The Bangladeshi population in the UK has a persistently higher stroke mortality, despite studies conducted in East London and Newcastle demonstrating that Bangladeshi adults tend to have a lower average mean systolic blood pressure than white Europeans [5]. However, Bangladeshi men tend to have a high prevalence of diabetes mellitus, smoking and physical inactivity, which could account for the increased stroke mortality. These observations may be relevant in Indians, Sri Lankans and Pakistanis, who also share geographical, physical, racial, genetic and ideological similarities.

So what do we know for sure about ethnic differences in populations? A geographical outlook is helpful in streamlining available evidence. The NOrthern MAnhattan Stroke Study (NOMASS) [6] was a population-based, case–control design to determine the prevalence and etiological fraction for stroke risk factors among whites, blacks and Caribbean Hispanics living in the same geographically defined community. Blacks and Caribbean Hispanics had a greater stroke etiological fraction for hypertension and diabetes than whites, a lower stroke etiological fraction for atrial fibrillation and coronary artery disease, and an equal etiological fraction for physical inactivity, thus elucidating the disproportionate burden of these modifiable risk factors among whites, blacks and Caribbean Hispanics, and suggesting that differential race–ethnicity prevention strategies may be needed [7].

Similarly, a cross-sectional sample from the National Health And Nutrition Examination Survey (NHANES) III [8] of the noninstitutionalized US stroke-surviving population found blacks to have a higher prevalence of hypertension, diabetes, peripheral vascular disease, C-reactive protein and inactivity, whereas white subjects had a higher prevalence of older age, myocardial infarction and lower high-density lipoprotein cholesterol. After adjustment, it was found that ethnicity and income were independent risk factors for stroke, whereas neither education nor insurance altered the ethnicity–stroke association. An obvious limitation of this study was the lack of inclusion of the higher-risk and higher-mortality associated hospitalized patients with stroke.

The Greater Cincinnati/Northern Kentucky Stroke Study (GCNKSS) was designed to measure incidence and trends of all strokes within a well-defined, large, biracial population [9]. This study also looked at stroke subtypes and found varying incidence rates according to stroke subclassification (using International Classification of Diseases [ICD]-9). For example, there was an excess burden of small-vessel stroke among blacks, in spite of similar prevalence of diabetes compared with whites, although cigarette smoking and hypertension were more prevalent in blacks. Contrary to the results of earlier studies, the GCNKSS also found a higher incidence of large-vessel and cardioembolic strokes among black populations, thus challenging earlier established concepts and begging for a need to better understand the underlying differences and the racial disparity of the ischemic stroke subtype.

Across the continent, events reveal even more exciting trends as data have been analyzed in first-generation immigrants, and particular paradoxical observations make compelling reading. The Leicestershire Stroke project showed a lower incidence of 30-day stroke mortality among South Asians compared with ethnic whites and, contrary to earlier beliefs, it was also noted that South Asians were less likely to be managed at home and had a higher hospitalization rate, thus dispelling the commonly held belief that South Asians prefer domiciliary treatment [10].

Does this mean that health districts should carry out local ethnic surveys on their population so as to streamline their efforts in tackling stroke morbidity and mortality? The South London Stroke Register follow-up study results reported an increased likelihood of survival among blacks compared with whites with a primary stroke, despite the fact that blacks were more likely to have a stroke, having a younger mean age of developing a stroke and lower socioeconomic status [11]. Quite similar to the US black population, the South London black community were less likely to have atrial fibrillation and more likely to have hypertension and diabetes mellitus compared with whites. It was felt that although the black population was at an increased risk of stroke, this was not translated into significant differences in survival. The reasons for this paradox remained unclear. To further confound matters, stroke registers between Barbados and South London were compared and it was found that the risk of stroke of a black Caribbean was much higher in the South London population and that the risk of a stroke for a black Caribbean in Barbados approached that of a white person in South London [12]. Thus, it can be seen that universal rules cannot be applied to different ethnic and racial groups [13], and migratory influences among populations may have an important role to play.

Similarly, Bangladeshi men, who have a higher prevalence of smoking, diabetes and sedentary lifestyles, are more likely to develop a stroke in comparison with their equivalent ‘high-risk’ white counterparts; however, another London study showed that they were also less likely to have their cholesterol levels checked and statin treatment initiated [14]. In comparison, a recent population survey of Gujaratis living in Sandwell (UK) and Navsari (India) demonstrated a higher prevalence of mean body mass indices, increased low-density lipoprotein (LDL) cholesterol and obesity among British Gujaratis, possibly accounting for higher cardiovascular disease risk among the migratory population [15].

In addition, a review of blood pressure studies in the South Asian population, in comparison with white people in the UK, showed higher blood pressures in Indians, slightly lower blood pressure in Pakistanis and much lower blood pressure in Bangladeshis, dispelling the traditional belief that all south Asians have higher blood pressures compared with their white counterparts [16]. Other risk factors, such as vitamin D deficiency, metabolic syndrome and chronic inflammation, are also under intense scrutiny [17]. Indeed, a unique risk factor exclusive to South Asians, highlighted by Charaborti and colleagues, was the effect of squatting and straining at stools, causing surges of early morning blood pressures and leading to an increased incidence of early morning stroke among this cohort [18]. Whether Indians and Bangladeshis who use a western-style commode have a lower risk of stroke remains unknown. This apparent heterogeneity in risk factors and emerging new concepts stress the need for more detailed studies in the future among ethnic groups.

What about the far Eastern population? Studies in China have shown an increased overall incidence and a clear north–south gradient of hemorrhagic strokes compared with the western population, although an in-depth analysis showed that serum cholesterol was higher in northern and urban areas, compared with rural southern populations, and there was a correspondingly higher proportion of ischemic stroke and a lower proportion of hemorrhagic stroke in urban and northern populations in China [19]. Compared with western populations, the Chinese have a lower level of serum cholesterol and body mass index, and have relatively more hemorrhagic strokes.

This not only explains the variance among Chinese populations, but also the differences between Chinese and western populations. In parallel with observations in other countries, stroke mortality has been declining in China, especially for men. Although no single cause was found, the reduction in stroke mortality appeared to be the result of better surveillance and improved medical care for stroke patients. While the decreasing stroke mortality is encouraging, the absolute number of patients with first-ever stroke is still escalating as the population ages. Whether this is due to increased prevalence of other cardiovascular risk factors and the adoption of a western diet and lifestyle remains under investigation. Indeed, baseline data collected in four Chinese population samples show contrasting findings both among Chinese samples and USA population samples for blood pressure, serum lipids, body mass index and smoking [20].

In conclusion, complex ethnic disparities exist in stroke incidence, risk factor profile and stroke mortality. The observed disparities are mainly based on the landmark studies from developed countries. Owing to the paucity of data from developing countries, understanding the ethnic differences in stroke in Eastern and Asian countries are very limited. Nonetheless, the available literature highlights the magnitude of the adverse relationship of stroke with black ethnicity at all level of the disease. Further attention is clearly needed in areas where available evidence is not conclusive or robust.