En Español | Plain Text Version | Skip Navigation

You are here:  Home > News and Events > Proceedings

Proceedings of the Stroke Disparities Advisory Panel Meeting

Skip secondary menu

National Institute of Neurological Disorders and Stroke

Proceedings of the Stroke Disparities Advisory Panel Meeting

November 7-8, 2002
Bethesda Marriott
Bethesda, Maryland

Table of Contents

EXECUTIVE SUMMARY
NINDS Stroke Disparities Advisory Panel Meeting
November 7-8, 2002
Bethesda, Maryland

INTRODUCTION

In 2001, the NINDS charged its Stroke Progress Review Group (Stroke PRG) with the task of setting overall priorities for stroke research. Because racial and ethnic disparities in the stroke burden present a critical research and treatment challenge that must be addressed, the Stroke PRG in its April 2002 report discussed the need to develop a research agenda on stroke disparities. In response, NINDS formed a Stroke Disparities Advisory Panel, which convened a preliminary meeting on June 11, 2002, to plan a full-scale workshop on future research directions in stroke disparities. The workshop took place on November 7-8, 2002, in Bethesda, Maryland. During this meeting, nine panels of experts were formed to discuss and consider recommendations for advancing research on stroke disparities.

BACKGROUND

Stroke ranks as the third leading cause of death and is the leading cause of serious, long-term disability in the United States . Racial and ethnic minority populations, in some age groups, have a higher relative risk of stroke death when compared with the U.S. non-Hispanic white population. For example, among African-Americans, the relative risk of stroke death is 4 times higher at ages 35-54, 3 times higher at ages 55-64, and almost 2 times higher at ages 65-74. The disparities in stroke mortality rates closely parallel racial and ethnic differences in the prevalence of stroke risk factors. For example, compared to white Americans, African-Americans have higher rates of hypertension, diabetes, smoking, obesity, and physical inactivity. These differences often persist even after statistical adjustments are made for differences in the socioeconomic status.

A number of possible mechanisms involved in stroke disparities have been hypothesized. They include a wide range of cultural and environmental factors, such as racial and ethnic variations in lifestyle, access to healthcare, quality of healthcare received, differences in health beliefs, religiosity, health literacy, adherence to prescribed therapy, stress, and exposures to environmental toxins. In addition, recent progress in stroke genetics has enabled exploration of possible racial and ethnic differences in genetic susceptibility to stroke or stroke risk factors. The panels reviewed the state of scientific evidence on the influence of environmental and genetic factors, and their interactions, on stroke incidence and stroke outcomes.

DISCUSSION

The discussions regarding current state of research on stroke disparities were conducted by nine panels, each focusing on one of the following topics:

  1. Temporal Trends in Disparities in Stroke Incidence, Mortality, and Quality of Life
  2. Disparities in Stroke Risk Factors and Mechanisms
  3. Access to Health Care/Quality of Care
  4. Genetics and Vascular Biology
  5. Design of Clinical Trials Addressing Stroke Disparities
  6. Bridges to the Community: Assuring Ethical Conduct of Studies and Data Integrity
  7. Primary and Secondary Stroke Prevention
  8. Acute Stroke Treatment
  9. Rehabilitation/Outcomes

Each panel summarized the state of current knowledge within their field of expertise and identified barriers and challenges that hamper our understanding of stroke disparities. The panels specified data needs and scientific standards of evidence necessary to design effective interventions and to develop reliable methods of measuring progress toward the elimination of stroke disparities. Several common themes have emerged across the panel discussions:

  1. Data needs

    The panels have consistently noted the urgent need to collect more data on racial and ethnic stroke disparities, going beyond the well-documented differences in stroke mortality. Racial and ethnic disparities in stroke mortality rates may reflect disparities in incidence rates or case-fatality rates, or both. However, stroke incidence and case fatality data from population-based studies that include substantial numbers of minorities are limited. Data regarding temporal trends in stroke incidence in minority populations are particularly important for measuring progress toward elimination of stroke disparities.

    Evidence to date has suggested possible racial and ethnic differences in the incidence of various stroke subtypes. For example, existing data suggest greater risk of lacunar and intracranial atherosclerotic stroke in African-Americans and Hispanics. Therefore, the panels have emphasized the need to collect further data on racial and ethnic differences in the incidence of different stroke subtypes and the relative contribution of different risk factors to increased rates of various stroke subtypes.

    There is strong and consistent evidence of low levels of implementation of established guidelines for primary and secondary stroke prevention by healthcare providers, and low levels of patient adherence to physician-prescribed medical therapy or changes in lifestyle. These shortcomings have greatly reduced the potential of existing evidence-based knowledge to reduce the stroke burden. However, little information is available on racial and ethnic differences in the quality of preventive care, including possible provider bias, and patient compliance as mechanisms contributing to the disparities in stroke burden.

    Numerous studies have documented racial and ethnic differences in medical care received for various conditions. For example, evidence shows that African-American cardiac patients are less likely than white American patients with the same condition and similar characteristics to receive diagnostic procedures, revascularization procedures, and thrombolytic therapy. In general, disparities in receipt of appropriate care remain after adjusting for factors known to affect care, such as age, sex, insurance status, co-morbidities, and disease severity. Despite the perception that there are racial disparities in acute stroke treatment, currently little is known about possible ethnic and racial differences in the treatment of acute stroke. For example, it is not known whether there are racial and ethnic differences in the use of thrombolytics for acute stroke treatment.

    Similarly, very little information exists on possible racial and ethnic differences in the rate of recovery from stroke, stroke recurrence, and functional outcomes and quality of life following stroke. Previous studies have shown that functional outcome after ischemic stroke is affected by many factors. Those most commonly reported include age, pre-stroke disability, severity of stroke and/or level of consciousness on presentation, presence or absence of urinary incontinence after stroke, previous stroke, diabetes or elevated serum glucose, cardiac disease, and degree of social support or marital status. However, in the majority of these studies, race was not considered as a variable potentially affecting post-stroke outcome.

  2. Challenges encountered in research on stroke disparities

    The panels consistently emphasized the need to improve and standardize race and ethnicity definitions and classifications. Definitions of race-ethnicity have been mandated by the government organizations, but may be inadequate and differ across available studies. These definitions were based on US census methods and do not take into account cultural distinctions, heterogeneity among race groups, ethnicity, heritage, and the effects of inter-marriage.

    There are major confounding factors when examining racial/ethnic disparities in the burden of stroke, including socioeconomic status, education, religion, cultural factors, dietary patterns, geographic region, and gender. These factors complicate study design and limit our ability to generalize results from one study to the entire U.S. population.

    Stroke is a heterogeneous disease and comprises a number of pathological conditions. The phenotypic characterization of stroke in research is a controversial issue that will require a diversity of approaches. While there is consensus that ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage should be addressed separately, an important question is whether there are subphenotypes or intermediate phenotypes that have a particular relevance to racial and ethnic differences in stroke risk.

    There are many operational challenges to conducting research in different racial and ethnic groups. The lack of legal protection against genetic discrimination in the United States , and general mistrust among minorities toward the medical and research establishment, are substantial impediments to the inclusion of racial and ethnic minorities in biomedical, especially genetic research. Participation in clinical trials is declining in numerous disease areas. Unfortunately, recruitment strategies have focused on problem-solving strategies rather than on proactive, specific strategies targeted to reach historically underserved populations.

RECOMMENDATIONS

Scientific Priorities

  • Epidemiological research

    Characterize temporal trends in racial and ethnic disparities for stroke mortality, incidence, and prevalence rates by age, gender, social class, social environment, and geography.

    Identify population attributable risks and interactions among known risk factors and emerging stroke risk factors in multicultural populations, accounting for differences in sex and socioeconomic status.

    Determine whether intermediate markers of stroke risk are valid markers in different racial and ethnic populations and potential targets for prevention trials.

  • Research on genetic and environmental determinants of stroke risk

    Provide characterization of variation in candidate genes and pathways within each racial and ethnic group, based on adequately powered studies within each racial and ethnic group, and with appropriate consideration of the potential for population stratification bias.

    Examine how genetic and environmental factors and their interactions contribute to the incidence of stroke risk factors and subtype-specific strokes.

    Study racial and ethnic differences in response to drugs used to treat acute stroke, as well as those used for primary and secondary prevention.

    Focus translational research on similarities and differences between racial and ethnic groups in important vascular biology mechanisms, including endothelial function, biomarkers of inflammation, coagulation/fibrinolysis, angiogenesis, and oxidative stress. Further characterization of these phenotypes between racial and ethnic groups must accompany genetic studies of these biological mechanisms.

  • Health services and patient management research

    Evaluate the independent contributions of factors relating access to quality care and stroke health disparities, such as health insurance status, acculturation, religious and spiritual factors, and healthcare provider attitudes. Research focused on prejudice and cultural competency on the part of health care professionals and their influence on access to care and stroke disparities should be undertaken. Investigate whether quality-of-care measurements and organizations can effectively ascertain the extent and possible causes of certain of the stroke-care disparities.

    Evaluate barriers to adherence to stroke prevention strategies that are specific to minority groups.

    Develop and test for effectiveness of novel cost-effective and practical stroke prevention programs for multicultural groups that can be adapted for use within communities.

    Collect data on access to acute stroke care among racial and ethnic subgroups. Evaluate the extent to which socioeconomic and cultural factors, level of public awareness, physician biases, and patient co-morbidities may influence acute stroke care.

    Using multidisciplinary research and considering a variety of sociological, environmental, and access variables, conduct research to learn how race affects access to and quality of stroke rehabilitation services. Investigate whether racial differences occur in the recovery trajectory and in optimal recovery after stroke; and determine whether disparities exist in other long-term sequelae of stroke and what factors influence these outcomes.

Research Methodology

Study diverse racial and ethnic groups in different geographic, social, and cultural environments. Cross-cultural studies or migration studies of racial/ethnic groups are useful scientific strategies to disentangle the role of race as a confounder reflecting both environmental and genetic differences. The effect of race/ethnicity on disease outcome should be examined after stratifying on numerous potential confounders, including measures of socioeconomic status, education, and health care access and utilization.

The need to increase minority representation in clinical trials remains critical. In addition to increasing sample size of minority groups, several strategies were suggested to increase the statistical power to characterize race as an effect modifier. For example, statistical power can be increased by the use of surrogate endpoints, such as a composite endpoint of clinical or silent (MRI) strokes. Another approach might be to expand the sample size to test race as an effect modifier, while minimizing the complexity of the study. In addition, the panels urged standardization of data elements on stroke risk factors and stroke outcomes, and data sharing across studies.

Resources Needed

The foundation toward characterizing the public burden of stroke it to develop a national surveillance system to establish and compare incidence and prevalence rates for stroke, both overall and by stroke subtype, for all major racial and ethnic subgroups. This system should include population-based studies in multiple well-characterized communities in different geographic regions throughout the United States . The surveillance system should employ common methodology and definitions.

Develop additional measures of race and ethnicity that can complement current methods of self-reporting and ensure consistency in reporting race across epidemiologic studies.

Strengthen the regulatory environment that protects against genetic discrimination of individuals and named populations.

Develop Stroke Prevention Centers of Excellence, to test interventions aiming to improve screening for risk factors, increase implementation of risk factor management guidelines by healthcare organizations and healthcare providers, and to support patient adherence. Such interventions should be designed by multidisciplinary teams, drawing on experience from diverse fields of research on healthcare delivery, healthcare outcomes, and health behavior.

A more proactive approach to minority recruitment, both investigator and patient, was recommended. Several strategies were suggested, including providing diversity training for investigators to increase their awareness of and sensitivity to cultural norms of target populations; creating a "special populations" advisory group to work with researchers; and making modifications in the NINDS guidelines to establish research requirements for community engagement.

Top


Panel 1
Temporal Trends in Disparities in Stroke Incidence, Mortality, and Quality of Life

Co-Chairs: Joseph P. Broderick, M.D., and George Howard, Dr.P.H.

Participants:
Ann M. Malarcher, Ph.D.
Lewis B. Morgenstern, M.D.
George Petty, M.D.
Philip Wolf, M.D.

STATEMENT OF THE PROBLEM

Although stroke is the third leading cause of death in the United States our current understanding of stroke burden among groups and over time is insufficient to provide the foundation needed to implement effective strategies to reduce stroke mortality and morbidity in the foreseeable future. Members of certain racial and ethnic groups such as African Americans bear a substantially greater share of the burden of stroke than do Caucasian Americans as measured by stroke mortality rates in this country. However stroke incidence data from population-based studies that include substantial numbers of minorities are limited as are data regarding temporal trends in stroke incidence in these minority populations. Few studies have performed assessments of quality of life following stroke in minority populations. Stroke prevalence data concerning racial and ethnic disparities are limited to data from a single question on a national survey asking whether a person has ever been told that they have had a stroke. Without the appropriate data strategies for eliminating racial disparities in stroke are unlikely to be effective and progress will be poorly measured.

  • Racial and ethnic designations in epidemiologic studies of strokes are usually obtained through self-reporting; however, such data are often missing in administrative databases. Methods of reporting race and ethnicity may vary between different epidemiologic studies of stroke.
  • Major confounding factors are found when examining racial and ethnic disparities in the burden of stroke, including socioeconomic status, education, cultural factors, geographic region, and gender. These factors complicate study design and limit our ability to generalize results from one study to the entire U.S. population.
  • Stroke is heterogeneous and comprises a number of pathological conditions; these conditions share the common manifestation of brain injury caused by disease of the cerebral vasculature. Information regarding racial and ethnic disparities is particularly lacking regarding incidence rates and mortality of specific stroke subtypes. Race and ethnic disparities in stroke may vary by major type of stroke (e.g., intracerebral hemorrhage versus ischemic stroke) and even by subtype of ischemic stroke (cardioembolic versus large-vessel atherosclerosis). Incidence studies thus must include stroke subtyping through brain imaging and other diagnostic tests.
  • Racial and ethnic disparities in stroke mortality rates may reflect disparities in incidence rates or case fatality rates or both. Incidence and case fatality data from population-based studies of biracial or multiracial populations are needed to address this issue.

CHALLENGES AND QUESTIONS/ BARRIERS

Lack of Data

Overall stroke mortality data are available for the various racial groups by age, gender, and geographic region, but they are limited regarding specific stroke subtypes. Incidence data by racial subgroup has been mostly limited to predominantly all-Caucasian, middle-class populations such as the Rochester Epidemiology Project or cohort studies such as the Framingham Heart Study, which may not adequately reflect incidence data from minority populations or other geographic regions.

The Greater Cincinnati/Northern Kentucky Stroke Study and the Northern Manhattan Stroke Study have reported stroke incidence rates in biracial and multi-ethnic populations at a single point in time, but temporal trends in stroke incidence rates in these diverse populations have yet to be reported. Several NIH-funded population-based studies are in progress, such as Multi-Ethnic Study of Atherosclerosis (MESA), Jackson Heart Study, Reasons for Geographic and Racial Differences in Stroke (REGARDS), and a study of non-Hispanic Caucasians and Hispanic Caucasians in Corpus Christi, Texas, but incidence rates from these studies have not yet been reported. Minimal data are available regarding incidence data for other minority populations such as Asian Americans and Native Americans. These population-based studies have published very limited information concerning measures of functional outcome and quality of life by racial subgroups.

Cohort studies such as the Atherosclerosis Risk in Communities study and the Cardiovascular Health Study are another source of incidence data and case fatality rates in different racial and ethnic subgroups, but they are not population-based studies. The advantage of the cohort studies are well-described profiles of risk factors in addition to subsequent case events. A disadvantage is the small number of cerebrovascular events, particularly for certain stroke subtypes.

Lack of Accurate Subtype-Specific Data

Published data in these population-based studies regarding incidence rates of stroke subtypes by racial and ethnic subgroups are very limited. For less common stroke subtypes such as subarachnoid hemorrhage, the study population needs to be relatively large in order to examine differences in racial and ethnic subgroups.

Inconsistency in Classification of Race and Ethnicity

Race is not a biological construct; it is a social one. There is also inconsistency is how race is classified between studies and in the presence of racial classification in certain databases (e.g., hospital discharge databases). Mortality for certain racial subgroups is substantially underestimated because of misreporting of race and ethnicity on the death certificate.

Need to Understand Cultural Contexts

Race and ethnicity are closely associated with socioeconomic status and cultural factors. Cultural contexts needed to understand differences by race and ethnicity and socioeconomic status include measures of discrimination, stress, acculturation, language barriers, access to healthcare, and others.

Access to Technology

Technology, such as computed tomography imaging and magnetic resonance imaging, can change detection and classification of stroke and stroke subtype and thus can affect measurements of temporal trends in the incidence rates of stroke overall and for subtypes. If access to this technology varies by race/ethnicity/socioeconomic status, measurements of incidence and prevalence rates of stroke also may be affected.

Poor Understanding of the Epidemiology of Outcome and Quality of Life Following Stroke

Epidemiology of outcome as characterized by cognitive and functional disability and quality of life following stroke is lacking. The identification of factors associated with positive functioning outcomes is the first step of providing interventions that may maintain or improve the quality of life for the stroke patient, as well as for caregivers, following stroke events.

RESEARCH AND SCIENTIFIC PRIORITIES

Priority 1:

Characterize temporal trends in racial and ethnic disparities for stroke mortality, incidence, and prevalence rates by age, gender, social class, social environment, and geography.

To characterize the public health burden of stroke, develop a national system to establish and compare incidence and prevalence rates for stroke, both overall and by stroke subtype for all major racial and ethnic subgroups. This surveillance system should include population-based studies in multiple well-characterized communities in different geographic regions throughout the United States. This system needs to provide detailed information on incidence rates, with strata defined by age, geographic region, and race and ethnicity. It should employ common methodology and definitions to the extent possible. These strata-specific incidence estimates will serve as the foundation for estimating the proportion of the well-known differences in stroke mortality rates between the strata that are attributable to differences in incidence, rather than case fatality, as well as the case-mix of stroke subtypes between the strata.

Of equal importance, this resource will provide the mechanism for prospectively tracking temporal changes in incidence, allowing for appropriate shifts in resources and research efforts in response to changes in the incidence rates (both overall and by stroke subtype), as well as shifts between the strata defined by demographic factors including racial and ethnic groups.

Priority 2:

Measure differences in stroke outcome and quality of life for different racial and ethnic subgroups.

A well-designed national system should also include cohorts of the various racial and ethic groups of stroke patients that can be prospectively followed to measure outcome and quality of life. Further methodologic research is needed regarding measurement of quality of life for stroke patients in general as well as the impact of stroke among diverse racial and ethnic subgroups. For example, racial and ethnic subgroups may place different values on various aspects of the quality of life following a stroke.

Information regarding functional outcome and quality of life from these cohorts will serve the secondary aim of establishing the magnitude and determinants of the public health burden associated with the post-stroke period for the racial and ethnic subgroups. Specifically, these cohorts can be used to estimate the costs associated with long-term treatment and the mortality, recurrence, and morbidity (including both recovery and cognitive decline) associated with the stroke event. For each of these domains, the cohort also can be used to establish the determinants that place stroke patients at risk for differential outcomes.

Finally, links between the national surveillance cohort and administrative databases (e.g., Medicare, Medicaid) can be investigated to provide determinants of cost-effectiveness and patterns of resource utilization on the national level.

Priority 3:

Find additional measures of race and ethnicity that can complement current methods of self-reporting and ensure consistency in reporting race across epidemiologic studies.

Race and ethnicity are social constructs that include multiple variables such as ancestry, migration, and acculturation. To compare stroke occurrence and outcome between diverse populations in various geographic regions, consistency of definitions and of reporting is critical. Self-reporting of race or ethnicity will likely remain the primary basis for comparison. However, all epidemiologic studies should indicate the number of cases for which racial and ethnic designation is missing.

Even within racial and ethnic groups, substantial variability may occur in ancestry and genetics as well as cultural environment. Additional information, such as ancestry and migration history, as well as genetic markers, will be useful for comparisons between similar racial or ethnic groups (by self-report) from different geographic regions.

RESOURCES NEEDED

  • Maintenance of current population-based studies of both Caucasian and non-Caucasian populations and the addition of other population studies (cohort, case control, etc.) for measurement of temporal trends in stroke incidence and prevalence rates as well as determinants of outcome in targeted racial and ethnic subgroups. In particular, additional research should focus on currently understudied minority populations.
  • The development of mechanisms for the linkage of information collected in the epidemiological studies, including the development of common methods and infrastructure for sharing stroke data by "outside" investigators. As part of this resource, consistent documentation, ideally using common definitions, needs to be developed for domains assessed in the studies.
  • The development of links with epidemiological databases from other sources (e.g., insurance companies and health maintenance organizations), including improved methods for clinical informatics.
  • Research approaches that ensure access to patient data while protecting patient confidentiality.
Top

 

Panel 2

Disparities in Stroke Risk Factors and Mechanisms

Co-Chairs: Ralph L. Sacco, M.S., M.D., Herman A. Taylor, M.D., and K.S. Lawrence Wong, M.D.

Participants:
Robert Adams, M.D.
Marco DiTullio, M.D.
Karen L. Furie, M.P.H., M.D.
James M. Galloway, M.D.
Edgar J. Kenton, M.D.
Stanley Tuhrim, M.D.

STATEMENT OF THE PROBLEM

Although stroke is the third leading cause of death in the United States , our current understanding of its etiology and variation among various racial and ethnic populations is not sufficient to develop the strategies needed to reduce stroke mortality and morbidity in the foreseeable future. Indeed, the lack of understanding of variations in risk factors and the associated mechanisms producing disparities in stroke incidence, mortality, and morbidity in various racial, cultural, and gender population subgroups represents a major gap in our knowledge. Some of these deficiencies have been noted in NINDS’ Report of the Stroke Progress Review Group, published in April 2002. Stroke is a heterogeneous disease. In addition to hemorrhagic stroke, there are different subtypes of ischemic stroke, and studies over the past decade have clearly identified factors associated with increased risk of both hemorrhagic and ischemic stroke. Some epidemiologic studies also have demonstrated racial and ethnic variation in stroke subtypes as well as these risk factors, which include a greater risk of lacunar and intracranial atherosclerotic stroke and a greater prevalence of hypertension and diabetes in African Americans and Hispanics than in nonminority populations.

In addition, recent work has shown that a significant proportion of stroke-prone individuals fail to receive appropriate therapy. Although evidence-based recommendations that address identified risk factors have been widely disseminated, stroke-prone women and minority populations, in particular, generally do not receive appropriate guidance for modifying even well defined, highly prevalent, and readily treatable factors. Thus, the greatest burden of risk for stroke is borne by underserved segments of the population—those least likely to benefit from advances in stroke prevention.

Currently, we do not have the data needed to address these disparities in stroke risk factors and associated stroke mechanisms or to develop an underlying approach to the treatment and modification within such subgroups of the population. The pursuit of these data and approaches will become increasingly essential as the 21st century progresses, as by 2030, those at risk for stroke will be more urban, poorer, and have lower educational levels than those at risk today and will comprise a major portion of the population, placing a tremendous burden on our public health and medical services. The first step needed to be better prepared to effectively prevent stroke in these population subgroups involves developing a more complete understanding of the inter-relationships between race and ethnicity, stroke risk factors, and stroke mechanisms.

CHALLENGES AND QUESTIONS/BARRIERS

Major questions and challenges center around the inadequacy of current definitions of race and ethnicity, the lack of minority-specific data on stroke risk factors and mechanisms, inadequate definitions and delineation of racial and ethnic, socioeconomic, and cultural confounding factors, and the lack of data regarding stroke outcome and recurrence risks.

Methodological Inadequacy of Definitions of Race/Ethnicity

Definitions of race and ethnicity have been mandated by government organizations, but they may be inadequate, and they differ across available studies. These definitions have been based on U.S. Census methods and do not take into account cultural distinctions, heterogeneity among racial groups, ethnicity, heritage, and the effects of intermarriage.

Lack of Minority-Specific Data

Although there are a few exceptions, most large-scale epidemiological studies of stroke risk factors and mechanisms focus primarily on majority populations. Further advances in the understanding of these risk factors and mechanisms as they pertain to minority populations will require additional data and the use of appropriate study designs to determine the incidence and outcomes of events and the prevalence and potency of risk factors. Studies to identify new risk factors and describe the incidence of stroke subtypes may provide new data that could help design stroke interventions. In addition, more specific data with regard to stroke subtype and mechanism are needed in order to better understand the differential effects of risk factors among and within racial and ethnic groups. This will require including individuals in studies with a broad representation of the factors, such as education and socio-economic status, which frequently confound issues of race and ethnicity.

Poorly Defined Racial and Ethnic, Cultural, and Socioeconomic Confounding Factors

Attempts to understand the effects of potential confounding factors, such as socioeconomic status, are themselves confounded by inadequate definitions and measures of discrimination, acculturation, language barriers, access to health care, effects of stress, the meaning of disease, and the possibility of health interventions within different cultures.

Inadequate Data Regarding Stroke Outcome and Recurrence Risk Among Minority Populations

Certain minority groups, most notably African-Americans and Hispanics appear to be at increased risk for stroke recurrence when compared with majority populations. The risk factors for stroke recurrence generally are poorly understood, and there is a paucity of racial- and ethnic-specific data in this area. The risk factors for subsequent stroke may differ from those for initial stroke, and, similarly, the factors affecting the disparity in recurrence risk may differ from those affecting initial stroke risk. A more profound understanding of these factors is crucial to addressing disparities in stroke recurrence. For example, little is known about the interaction of cognitive and functional disability following stroke and race and ethnicity as related to the risk of recurrence or other quality of life issues.

Endemic Barriers to Identifying Stroke Risk Factor Disparities

Current barriers, which need to be overcome in order to identify stroke risk factor and mechanism disparities, include:

  • A lack of uniformity in defining population subgroups and specific risk factors/mechanisms.
  • A lack of adequately powered studies to control for confounding variables in multicultural populations.
  • The need to identify culturally sensitive and ethical recruitment techniques to ensure adequate numbers of diverse subjects for such studies.
  • The need to overcome case ascertainment bias.
  • The presence of geographic and social barriers to accessing isolated populations.
  • The presence of cultural barriers to the acceptance of clinical research.
  • The presence of legislative and political obstacles to conducting research— particularly genetic studies—in disparate populations.

RESEARCH AND SCIENTIFIC PRIORITIES

The official United States multicultural classification scheme divides the population according to self-reported race and "Hispanic" versus "non-Hispanic" ethnicity, a scheme rife with misclassification problems and one that fails to account for the fact that genetic and cultural differences within these groups exist based on geographic origins, generations spent in the United States, and the potential for multiple racial and ethnic contributions through intermarriage. Studies examining population stratification in genetics suggest that inter-racial differences based on the current schemes may actually be quite small. It is critical that innovative definitions of subgroups incorporate cultural, gender, and socioeconomic factors in addition to race and ethnicity and classify groups based on susceptibility to, and risk from, factors that affect the incidence of stroke.

Priority 1:

Identify population attributable risks and interactions among known risk factors and emerging stroke risk factors in multicultural populations, accounting for differences in sex and socioeconomic status.

The prevalence of individual biological and behavioral risk factors is expected to vary in different racial and ethnic groups and differs by age, sex, and socioeconomic status. In addition, a specific risk factor may contribute differentially to stroke risk from population to population. The population attributable risk takes account of both the prevalence and relative risk of individual factors and is therefore a useful means of determining the extent to which an individual risk factor impacts a population. It is important to examine subgroup-specific susceptibility and unique interactions between risk factors. Studies will need to determine whether the prevalence of the condition and control of the condition vary significantly by race and ethnicity and account for differences in stroke incidence. This information is critical for designing epidemiological studies and clinical stroke prevention trials, as well as for focusing public health resources appropriately.

Priority 2:

Examine how genetic and environmental factors and their interactions contribute to the incidence of stroke risk factors and subtype-specific strokes.

The extent to which genetic susceptibility, as opposed to differences in lifestyle based on cultural mores and socioeconomic status, affects the incidence of risk factors and strokes due to specific mechanisms remains unclear. These factors are not mutually exclusive and, indeed, may have additive or synergistic effects on stroke risk. The explanations for observed patterns of cerebrovascular disease have yet to be elucidated. For example, the predominantly intracranial atherosclerosis observed in Asian populations, once believed to be genetically mediated, has been shown to evolve into the more common extracranial pattern seen in Caucasian Westerners with the alteration of environmental factors, but the underlying vascular biology that results in such changes is poorly understood. Other factors may predispose to other stroke mechanisms such as cardio-embolism or the development of lipohyalinosis in specific populations. It is unknown to what extent genetic regulation of factors such as endothelial function, immune response, or thrombosis could increase or decrease risk in a population independently or through amplification of conventional risk factors. Understanding these complex interactions will afford opportunities for screening and early intervention in high-risk populations.

Priority 3:

Determine whether intermediate markers of stroke risk are valid markers in different racial and ethnic populations and potential targets for modification trials.

Identification of preclinical markers of stroke risk and intermediate phenotypes is extremely important because it allows for the early identification of a high-risk population and provides a surrogate target for risk factor modification. In recent years, several markers of risk have emerged, but it is unclear whether they convey the same prognostic significance in all groups. Because mechanism of stroke may differ across racial and ethnic populations, markers that are gaining widespread clinical application may not be generalizable to all populations. For example, risk mediated through novel pathways (e.g., eNOS, vWf, fibrinogen, CRP, homocysteine, lipoprotein (a), IMT, LVH, among others) may be entirely missed by screening strategies developed through research on or for use in a predominantly non-Hispanic Caucasian population.

Priority 4:

Determine the risks, determinants, and mechanisms of adverse outcomes after stroke, including mortality, recurrence, dementia, disability, and quality of life in different racial and ethnic populations.

Outcomes after stroke may vary by racial and ethnic group and may be influenced by the prevalence and control of risk factors, stroke subtype, and social and behavioral factors affecting adherence. In addition, it is important to examine the contributions of the primary health care provider and the family caregiver to the determination of outcome. Further studies are needed to specifically address the determinants of outcomes after stroke in order to design better stroke outcome modification programs.

RESOURCES NEEDED

Assuming the goal is to better understand and address racial and ethnic disparities in stroke, especially with respect to risk factors and stroke mechanisms, the following are the resource goals that should be pursued and the resources that will be needed:

The overall resource goals are:

  • To adopt common definitions of terms across studies.
  • To stimulate and support new research to address the stated priorities.
  • To use existing data more effectively by creating a means to access and incorporate diverse sources of information, ranging from administrative data and high-level prospectively acquired data from NINDS, as well as other data sponsored by other granting agencies.
  • To develop innovative methods of pooling data.
  • To eliminate misclassification bias due to inadequate access to diagnostic testing/subspecialty expertise.
  • To expand the interactions among existing programs and facilitate communication across governmental and nonprofit organizations.

  1. Enhanced Data Collection

    Enhanced data collection will require the following:

    • Methods to standardize the operational definitions of race, putative risk factors, and mechanisms; clear definition of risks and mechanisms in future studies to enable data pooling.
    • Data collection from ongoing clinical studies, including randomized trials in diverse racial and ethnic populations.
    • New population-based data collection efforts in various U.S. regions targeting certain racial and ethnic populations.
    • Linked data collection from other administrative data sources. Reliance on investigator-driven special effort in select populations might need to be balanced by broad-scale linking of data from other sources. For example, larger numbers of representative patients can be analyzed through a concerted effort to identify and link disparate and diverse sources of data, including health maintenance organizations, insurance companies, registries, Centers for Medicaid and Medicare Services, Centers for Disease Control and Prevention (CDC), National Heart, Lung, and Blood Institute, Department of Veterans Affairs, Indian Health Service, and state-based data, where available. Attempts should be made to use sources from industry such as HEDIS indicators reported from managed care plans to assess possible differences in racial/ethnic application of stroke-pertinent physician/care plan behaviors (e.g., counseling and screening).
    • Use of data sources not yet exploited, including banked serum and DNA from prior clinical studies (e.g., measuring C-reactive protein or homocysteine in frozen samples from earlier studies with sufficient representation of different racial and ethnic groups).

  2. Centralized Data Collection and Analyses

    To move forward in accomplishing some of the research priorities and resource goals, it may be helpful to develop a centralized, national database for the consolidation of data sources and pooled analyses. This resource could accumulate data from various sources across agencies and organizations to form a data clearinghouse that could be accessed by various groups. This effort will require ongoing infrastructure support.

  3. Expanded Access to Diagnostic Facilities

    Widespread availability and access to diagnostic testing and subspecialty care will be needed to ensure the accurate classification of risk factors and stroke mechanisms in lower socioeconomic populations, for which the lack of availability of certain testing may lead to misclassification and inadequate diagnoses:

    1. Use of noninvasive vascular testing, magnetic resonance imaging and magnetic resonance angiography, and echocardiography, including transesophageal echocardiography, which are often needed to reliably classify stroke mechanisms.
    2. Improved access to subspecialty consultation.
    3. Uniform screening of blood pressure, cholesterol, glucose, weight, and lifestyle factors to identify subjects with undiagnosed risk factors.

  4. Stroke Prevention Research Centers on Racial and Ethnic Disparities

    To address these research priorities and help fill the knowledge gaps regarding racial and ethnic disparities in stroke risk factors and mechanisms, we strongly encourage NINDS to establish specific centers for stroke prevention research. These centers would focus on racial and ethnic disparities and stroke prevention and could include various study designs (e.g., epidemiology, mechanism, pilot trials, basic science, genetic studies) and encompass the range of stroke experience, including stroke precursors and persons at risk, as well as stroke survivors. Just as specific research centers have been established for acute stroke translational research (e.g., SPOTRIAS) and other statewide data collection stroke registries have been initiated (e.g., CDC Coverdell centers), the establishment of stroke prevention research centers would help focus efforts on these issues.

    Such centers would include diverse racial and ethnic and cultural groups and could share definitions of terms and systems and approaches to facilitate the recruitment of under-represented groups. They would facilitate interaction among neurologists, cardiologists, primary care providers, behaviorists, epidemiologists, statisticians, geneticists, and those working in other appropriate disciplines to develop the most effective study designs for addressing disparities in risk factors and mechanisms. The centers should include links to local community hospitals in order to evaluate stroke prevention activities beyond the academic center, and partnerships with other groups, including other federal and state entities as well as nonprofit organizations such as the American Stroke Association and the National Stroke Association, would be encouraged.

Top


Panel 3

Access to Care/ Quality of Care

Co-Chairs: Yvonne Harris M.P.A., and Lewis B. Morgenstern M.D.

Participants:
Carol Allen Ph.D., R.N.
Hal Barron M.D.
Gloria Bonner Ph.D., R.N.
Edward Cooper M.D.
Cheryl Easley Ph.D., R.N.
Moises Gaviria M.D.
Karen Parko M.D.
Beth Quill M.P.H.

STATEMENT OF THE PROBLEM

Although access to health care can be thought of as an absence of barriers to obtaining needed health care services—that is, having adequate health insurance and a routine health care provider and health care facility—several other components related to access may be related to the problem of health care disparities among different populations and may increase the stroke burden for certain individuals. These components include socioeconomic status; acculturation, faith, and religion; social determinants of health (including fatalism, optimism, depression); and geographic isolation. A large body of literature shows that members of racial and ethnic minority groups frequently have poor access to medical care and receive lower quality of care than do nonminorities. At the same time, members of minority populations are less likely to have health insurance, have more difficulty accessing health care services, and have fewer provider options than nonminorities, even when they have commensurate insurance and when other factors related to health care access are identical.

The reasons for these disparities are complex, not well defined, and difficult to quantify, but they may reflect a variety of health-related risk factors; differences in access to medical care; and socioeconomic, social and cultural, psychological, and individual patient and provider differences. Access to care as a clear causal explanation for health disparities in stroke remains to be demonstrated, as intervention studies have not yet been conducted to determine whether eliminating access to care deficits will reduce stroke disparities in the United States .

CHALLENGES AND QUESTIONS

  • Precisely define and prioritize barriers to access to quality health care services for minority populations, and increase research in this area.
  • Formulate and test interventions to improve access to quality health care for minority populations.
  • Apply the most successful and cost-effective interventions, and monitor the overall results against an established baseline for access.

It is important to define the aspects of insurance coverage, both public and private, including prescription drug coverage, that lead to the use of the best stroke prevention and acute stroke practices. Increasing the numbers and improving the location of minority health care professionals are also priorities. Additional consideration and study are needed to determine the impact of increasing the number of minority investigators, as well as minority patients in underserved communities, because the needs and effects may vary for different racial and ethnic groups. Stroke training also needs to be provided to all primary care providers and neurologists. The process of separating out socioeconomic factors and geographic location from access to care will help define the role of race and ethnicity in the area of receiving quality care for stroke, and recognizing acculturation and the stress that accompanies it as a modifier of health care access is critical. The role of faith, religiosity, and social determinants of health require further study to determine their interaction with stroke occurrence and outcome in various populations. It is also important to study emerging populations, including Asian Americans, Arab Americans, and Native Americans, while understanding that not all minority populations have the same access to care issues. Recruiting other federal agencies, industry, and other organizations to partner with NINDS to fund research in access to care and quality care for stroke prevention and acute stroke therapy is likely to enhance opportunities for reducing disparities. In addition, improving the delivery of acute stroke therapy to minority populations is likely to help reduce disparities.

BARRIERS

  • Economic Barriers: Financial resources and insurance are insufficient.
  • Service Barriers: Stroke services are lacking, not accessible, or not available.
  • Social and Cultural Barriers: These barriers include language barriers, lack of transportation, lack of awareness of services, service fragmentation, and under-representation of minorities among service providers.
  • Individual Barriers: Individual barriers include bias, discrimination, prejudice, and stereotyping practices at the provider, institutional, patient, and health system levels.
  • Psychological Barriers: These include distrust generated by individual and institutional discrimination and by cultural insensitivity.

A common misconception is that having insurance guarantees access to good care; the reality is that having incomplete insurance is a barrier to accessing quality stroke care and that having no insurance is, of course, a distinct barrier. For example, patients without prescription drug coverage are likely to be less compliant with stroke prevention therapies. Not having a routine health care provider is a common barrier to stroke prevention in underserved areas. Poor stroke training among primary care providers impedes access, as does insufficient training for health care professionals regarding culturally sensitive ways to foster trust. Similarly, few community outreach programs focus on encouraging trust in the medical encounter, a particular problem with acute stroke therapy, which includes an urgent consent process that necessitates such trust. Even within minority groups that have achieved generally high socioeconomic status, psychological stress resulting from racial discrimination is likely to be evident, and this is also true on an institutional level.

Acculturation and the related stress negatively impact health behaviors that are needed for stroke prevention and acute stroke therapy. The role of belief systems, including fatalism and its effects on stroke health behaviors, must be addressed, as must religious practices, which may have any number of effects on health-seeking behavior. Lack of information regarding stroke burden in emerging populations is a major obstacle, and problems involving language, prejudice, and a lack of minority physicians in immigrant groups represent notable barriers. In addition, industry's focus on targeting large groups at risk for stroke may lead to the neglect of minority populations that may benefit the most from stroke prevention services and acute stroke therapy.

RESEARCH AND SCIENTIFIC PRIORITIES

  • Descriptive epidemiologic studies are needed to evaluate the independent contributions of factors relating access to quality care and stroke health disparities. A specific focus on research using intragroup and intergroup comparisons is suggested.
  • Research focused on prejudice and cultural competency on the part of health care professionals and their influence on access to care and stroke disparities should be undertaken. At the same time improving trust among all populations in medical care and medical research is a critically important research topic.
  • Multidisciplinary research efforts are needed that combine the efforts of health services researchers, stroke neurologists and other stroke specialists, generalists, emergency medicine personnel, behavioral scientists, and representatives from the populations studied. Population-based studies are the method of choice.

Insurance

The independent contribution of insurance status separated from socioeconomic status in health care disparities requires continued research, and studies regarding insurance and health disparities among subgroups are needed, as are aggregate studies. Studies of stroke across private insurance plans and publicly funded care, including care provided in managed care systems, outside network systems, and private provider groups, are needed. The method employed should utilize community-based stroke research that explores the impact of community and public health resources on health outcomes. Studying the influence of continuity of care on stroke prevention is critical.

Acculturation

Studies should be undertaken to investigate the different acculturation patterns among minorities in the United States regarding stroke prevention and treatment, keeping in mind the heterogeneity of the different subgroups even within one minority population. Studies are warranted regarding the belief systems among the different racial and ethnic groups to determine their role in health behaviors related to stroke. Studies are needed that test the effectiveness of culturally sensitive interventions in closing the gap between research and health outcomes. Rehabilitation may be a prime focus for studying the role of acculturation and health disparities, as compliance with prescribed therapies for stroke prevention may be influenced by acculturation and acculturative stress. More studies are needed comparing stroke risk in recent immigrants with minority populations that have been in the United States for multiple generations.

Faith, Religion and Social Determinants of Health

Studies on religion and spirituality and stroke are needed that pay careful attention to social and behavioral mechanisms through which religion or spirituality influences health behavior. The relationship between religion and spirituality and the incidence of stroke, stroke risk factors, and lifestyle factors associated with stroke needs to be studied, as does the effect of religion and spirituality on the use of traditional or alternative sources of health care. Religious and spiritual factors that negatively impact health status and behaviors warrant further attention, and determining whether the integration of religious and spiritual issues into the delivery of primary care services would be a valuable approach. Exploring the efficacy of religious organizations as instruments of health promotion, health education, and the delivery of community-based health care services is another important focus of research.

Race and Socioeconomic Status

Better definition of race and ethnicity and socioeconomic status as constructs for research is needed, including consideration of the variations within racial and ethnic groups and the growing category of mixed-race groups. Race may be defined as a biological category for genetic studies or as a social and political category when studying the effect of exposure to individual and institutional racism on the risk of stroke and access to care and quality of care. Ethnicity studies should be used to understand the impact of cultural, lifestyle, and belief systems on risky or health-promoting behavior. Discrimination, skin color, preference in body size, diet, attitudes toward health care professionals, and use of alternative health care and their effects on access to care and stroke disparities are important topics for investigation. Additional studies investigating poverty are needed to relate its effects as a determinant of access to quality care. Poverty should also be studied to determine its impact on lifestyle, how it affects access to recreational facilities, and its contributions to stress and level of education and health literacy, as well as its relationship to emergency medical system response. Investigations of the interactions between race and ethnicity and gender also should be undertaken.

Emerging Populations and Native Americans

More information is needed on the current health care utilization patterns of racial and ethnic minorities for acute stroke treatment and stroke prevention, including knowledge, attitudes, and beliefs about acute stroke and stroke prevention among these groups. The risk of the disease must be clearly delineated in these populations, including the risk that is attributable to limitations in acculturation and access to care. Distinctions must be made between issues that are common to more recently immigrated populations and those inherent to stable United States populations. Emerging populations and Native American populations should be targeted for clinical research for prevention strategies for stroke, as they are currently excluded from such trials by their geographic isolation and their misunderstanding of the goals of such investigations. The communities involved should be queried regarding their willingness to participate in genetic studies, and studies should be encouraged that occur in communities outside of academic medical center communities. The differences in perspectives and recommended treatments between traditional medicine and Western-trained health care providers should be evaluated to understand how the two can interface.

Acute Stroke

More minorities should be recruited to help design, conduct, and participate in acute stroke studies in order to learn more about the value of acute stroke therapies in minority populations, and additional acute stroke care projects involving minority patients and minority health professional organizations should be encouraged. Research is needed to clarify whether discrimination occurs for delivery of acute stroke therapy. Further study is needed to determine the value of a stroke toolbox for health care professionals that includes a module regarding minorities and whether it should be included in such research. An important study focus is the expansion of Internet resources for management of acute stroke. Better studies on intravenous recombinant tissue plasminogen activator efficacy and risk in underserved groups are needed. In addition, information dissemination regarding stroke warning signs, the importance of calling for emergency assistance, and the availability of effective therapy for hard-to-reach minorities is a critical study issue.

Industry and Other Funding Agencies

Research is needed to determine whether industry-sponsored activities improve access to care and reduce stroke health disparities. It is important to determine whether the resources that industry uses for community activities are properly targeted. NINDS should collaborate with other federal funding agencies to facilitate mechanisms to perform these studies.

Multidisciplinary Collaborations

Studies investigating access to care and quality of care in minority populations should employ multidisciplinary investigative teams in order to maximize scientific benefit in the design, implementation, and analysis of data.

RESOURCES NEEDED

  • Funding is much needed to encourage research on issues related to access to quality care and minority populations in the United States . RFA/PA mechanisms to stimulate observational, epidemiologic, intervention, and outcomes research in this area are necessary. It is critical that funding agencies pool efforts to establish appropriate funding mechanisms.
  • Funding research that monitors the quality of care and that helps facilitate improvement in access to quality care for stroke in underserved populations would be optimal. Cost-effectiveness studies and cost-effectiveness analyses in stroke trials should be incorporated into such research.
  • A special emphasis review panel should be considered to review grant applications in stroke health disparities.
  • Additional resources are needed for community outreach efforts that will increase our understanding of the barriers to and facilitators for care within minority populations. Focus groups that include health care workers, members of the public, and local health advocates would be helpful in designing funding targets for improving access to health care.
  • Efforts to increase representation of minorities in the field of stroke research may help further our understanding of how to help prevent and provide the best treatment for stroke in minority patients.
  • Providing information about acute stroke therapy and stroke prevention in native languages is important in working to reduce disease burden in emerging populations. Web-based information, particularly information provided in an interactive format, would help improve health care access by helping to demystify the health care system and foster trust.
Top


Panel 4

Racial and Ethnic Disparities in Stroke: Genetics and Vascular Biology

Co-Chairs: Steven J. Kittner, M.D., M.P.H., Gary H. Gibbons, M.D., and Joseph P. Broderick, M.D.

Participants:
David Hess, M.D., M.S.
Richard Levy, Ph.D.
James F. Meschia, M.D.
David J. Pinsky, M.D.
Charles N. Rotimi, Ph.D.
Bradford B. Worrall, M.D., M.Sc.

STATEMENT OF THE PROBLEM

The impact of race/ ethnicity on health status results from a complex interaction of cultural, social, economic, and biological factors associated with racial categories. There are no distinct biological "races," but at the societal level, the concept of race has strong cultural, political and economic meaning. The operational construct of race becomes relevant in our attempts to understand disease distribution within and between populations. Race, although lacking consistent biological definition, has real biological consequences that are manifested at the physiologic level. For these reasons, racial categories are important descriptors in biomedical research. Ethnicity may be a preferable term that incorporates nongenetic factors including economic, social, religious, linguistic, dietary and other variables to identify individuals as belonging to a group, while allowing for potential differences in gene frequencies and for changes over time in both gene frequencies and nongenetic factors.

As one examines human genetic variation at a population level, there are geographical differences in the frequency of some genetic markers. Individual genetic make up reflects ancestral geographic origin, migration patterns, and population admixture or isolation. Genetic variability among member of a particular race/ ethnic group likely has a greater health impact than differences across these groups. In relationship to stroke, it is likely that differences in recognized and unrecognized environmental factors play a dominant role in observed disparities in risk across different race/ethnic groups. Thus, the medical and research communities must avoid an overly simplified concept of race as biological and the naPve assumption that differences across ethnic groups are genetically based.

Nonetheless, there are known biological determinants of stroke risk that differ across race/ ethnic groups. Higher frequencies of genetic variants of the human beta A globin gene in people of Mediterranean or African origin influence cerebrovascular risk in these populations. The causative genetic mutations are different in different ethnic populations for two hemorrhagic stroke syndromes (familial cerebral cavernous malformations and hereditary amyloid angiopathy). These examples underscore the potential impact of genetic factors on stroke risk in different ethnic groups. As one moves from disease causing genetic variants (monogenetic traits) to risk modifying or susceptibility genes (polygenic or complex traits), the difficulty of identifying the relevant genes increases. Yet, methods are currently available for studying stroke as a complex genetic trait and the potential public health importance of identifying susceptibility genes makes this an important scientific priority.

The risk of early-onset stroke death is increased among African-Americans, American Indians/Alaskan natives, Asian/Pacific Islanders, and some Hispanic populations. Among African-Americans, a major part of this increased risk is due to a higher prevalence of major established vascular risk factors, such as hypertension and diabetes mellitus, at earlier ages and a higher prevalence of multiple vascular risk factors. There may also be a genetic contribution to the disparity in stroke risk since there are differences between racial/ethnic groups in the prevalence of many candidate genes relevant to stroke. Even candidate genes with similar prevalence between racial/ethnic groups may contribute to the disparity in stroke due to gene-gene and gene-environment interactions.

We need improved strategies for stroke prevention, treatment, and recovery, and these strategies need to be established as effective for all segments of our society. In order to accomplish this goal, we must understand the underlying vascular biology, including the contribution of genetic factors, environmental factors, and their interplay. Diversity of genetic background and environmental circumstances provide opportunities for understanding the pathogenesis of stroke that are not present in homogeneous populations. Furthermore, candidate gene association studies that fail to properly account for race/ethnicity and subgroups with different genetic ancestry can have biased results, that obscure potentially important associations or mislead investigators with results due only to the genetic structure of the study population.

CHALLENGES AND QUESTIONS

The delineation of the role of genetic and environmental factors in observed racial and ethnic disparities in cerebrovascular disease, and in stroke risk generally, is a major challenge for stroke researchers. The genetic and environmental information conveyed by racial and ethnic groupings is correlated and confounded. Thus, it is critical to avoid misinterpretation of racial and ethnic differences as synonymous with genetic differences. Two challenges to understanding the complex web of genetic and environmental determinants of stroke are difficulties inherent in characterizing race and ethnicity and environmental factors in research studies.

Self-reported race is a proxy for social, cultural, and environmental factors that can predispose an individual to stroke, factors that would be otherwise difficult to measure. Current methods for classifying race and ethnicity in genetic studies of stroke are inadequate for this purpose. For example, consider the particular issues of ethno-history and population admixture that may complicate the generalization of results of studies of Mexican-Americans in San Antonio, Puerto-Ricans in New York, and descendents of African slaves from Guatemala to individuals who may fall under the general rubric of "Hispanic." Similar difficulties could occur in the aggregation of study results regarding Japanese and Pacific Islanders. Thus, self-identification criteria as employed by the Census are not adequate for capturing the social construct of race and ethnicity.

In addition, self-reported race inadequately captures differences between subpopulations in the frequencies of genetic markers. When case and control samples are drawn from different subpopulations, allele frequencies will tend to differ among the subpopulations for most randomly chosen loci. This can lead to spurious or biased estimates of allele-disease associations, because a disease that has a higher incidence in a given subpopulation will be associated with any alleles that have a higher frequency in that subpopulation. Several approaches are available for addressing this issue, including using the well-established transmission/disequilibrium test (TDT), which requires family controls, detailed information on familial ancestry, or the relatively new technique of population substructure analyses. To date, these approaches are infrequently used in stroke research.

The limitations of environmental exposure assessment also hamper the study of the genetic component of racial and ethnic differences, and established vascular risk factors are imperfectly measured. For example, the measurement of a person’s blood pressure at a single point in time underestimates the effect of this factor on stroke risk, even in homogeneous populations. This limitation is accentuated in comparisons that are made between racial and ethnic groups, for which there can be important differences in the age of onset of hypertension as well as in the diurnal variation in blood pressure. In addition, important social and cultural correlates of race and ethnicity are often not measured or are poorly measured. These nontraditional epidemiological variables include health beliefs and participation in medical care, access to quality medical care, and differential physician behavior. Given these difficulties, it is important to avoid the simplistic interpretation that group differences that persist after adjusting for other measured social and biological stroke risk factors are innate differences.

The phenotypic characterization of stroke for genetic studies is a controversial issue that likely will require a diversity of approaches. Although there is consensus that ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage should be addressed separately, an important question remains regarding whether there are subphenotypes or intermediate phenotypes that have a particular relevance to racial and ethnic differences in stroke risk. For example, ischemic stroke due to intracranial versus extracranial atherosclerosis and small subcortical strokes may be informative subphenotypes for ischemic stroke, and cortical versus subcortical intracerebral hemorrhage may be useful subphenotypes for intracerebral hemorrhage. Similarly, magnetic resonance imaging-defined infarcts, cerebral white matter disease, and endothelial dysfunction may be useful intermediate phenotypes for investigating racial and ethnic differences in stroke risk.

Racial and ethnic differences in response to commonly used medications to prevent and treat stroke are not sufficiently understood.