Differences between groups were significant by the log-rank test (P = .002).
eTable 1. Baseline Characteristics of U.S. Patients in the POINT Trial, Stratified by Missing Data on Race/Ethnicity
eTable 2. Association between Black Race and Recurrent Ischemic Stroke among U.S. Patients in the POINT Trial, Stratified by Subgroups
Customize your JAMA Network experience by selecting one or more topics from the list below.
Kamel H, Zhang C, Kleindorfer DO, et al. Association of Black Race With Early Recurrence After Minor Ischemic Stroke or Transient Ischemic Attack: Secondary Analysis of the POINT Randomized Clinical Trial. JAMA Neurol. 2020;77(5):601–605. doi:10.1001/jamaneurol.2020.0010
Do black patients in the United States face a higher risk of early ischemic stroke recurrence?
In this cohort study of 4044 US participants in the Platelet-Oriented Inhibition in New Transient Ischemic Attack or Minor Ischemic Stroke trial, 918 (22.7%) were black. In an adjusted Cox model, black race was associated with a higher risk of recurrence compared with white race.
Among US participants in a randomized clinical trial focused on secondary prevention after minor ischemic stroke or transient ischemic attack, black individuals faced a higher risk of early ischemic stroke recurrence, even after adjustment for demographics, comorbidities, and medication adherence.
Stroke incidence is higher among black than white individuals in the United States. It is unclear whether black individuals have a higher risk of stroke recurrence after a minor ischemic stroke or transient ischemic attack (TIA), a high-risk setting in which focused preventive efforts can be effective.
To examine the association between black race and early ischemic stroke recurrence.
Design, Setting, and Participants
This cohort study analyzed data from the Platelet Oriented Inhibition in New TIA and Minor Ischemic Stroke (POINT) trial conducted at 269 sites from May 28, 2010, to December 19, 2017. The trial enrolled 4881 adults within 12 hours of onset of a minor ischemic stroke (National Institutes of Health Stroke Scale score, 0-3) or high-risk TIA (ABCD2 score, ≥4). For this analysis, we excluded 598 patients enrolled outside the United States and 239 US patients with missing race/ethnicity data.
Main Outcomes and Measures
The primary outcome for this analysis was ischemic stroke within 90 days after randomization. Covariates included age, sex, Hispanic ethnicity, study assignment to take clopidogrel vs placebo, index stroke vs TIA, vascular risk factors, statin use, study drug adherence, and index event etiological subtype.
Among 4044 patients included in the analysis, 918 (22.7%) were black. In an adjusted Cox model, black race was associated with a higher risk of recurrence compared with white race (hazard ratio, 1.6; 95% CI, 1.1-2.3). Findings were similar in subgroup analyses and in analyses limited to sites that enrolled black patients.
Conclusions and Relevance
Among US participants in the POINT trial, black individuals faced a higher risk of early stroke recurrence after a minor ischemic stroke or TIA. Our findings support research into black-white racial differences in the underlying mechanisms of recurrent stroke. In the meantime, extra effort should be made to ensure that black patients have access to proven secondary prevention measures.
clinicaltrials.gov Identifier: NCT00991029
Quiz Ref IDBlack US residents face a substantially higher risk of ischemic stroke than white individuals.1 Eliminating this disparity would prevent approximately 80 000 strokes among black adults in the United States each year.2 While studies have consistently found a black-white racial disparity in the risk of incident ischemic stroke,1-4 there are conflicting reports about whether black individuals face a higher risk of recurrent stroke.5-9 Few data exist on the risk of recurrent stroke in the first few months after an incident ischemic stroke or transient ischemic attack (TIA), when the risk of recurrent stroke is highest. These knowledge gaps are significant because most initial strokes are relatively mild10 and provide an opportunity to institute evidence-based treatments to prevent more disabling recurrent strokes.11 Therefore, we examined the association between black race and the risk of early stroke recurrence among US participants in the Platelet Oriented Inhibition in New TIA and Minor Ischemic Stroke (POINT) trial.
POINT was a randomized clinical trial that compared the combination of aspirin and clopidogrel vs aspirin alone for preventing major ischemic events after minor ischemic stroke or high-risk TIA.12 A total of 4881 patients were enrolled from May 28, 2010, through December 19, 2017, at 269 sites in 10 countries. Patients were followed up for 90 days. The trial was approved by the ethics committee at each participating site and all participants provided written consent. The trial was funded by the National Institute of Neurological Disorders and Stroke (NINDS), which also provided the deidentified, publicly available data used in this post hoc analysis. The Weill Cornell Medicine institutional review board approved this analysis.
The POINT trial enrolled adults 18 years or older who could undergo randomization within 12 hours after the onset of a minor ischemic stroke (National Institutes of Health Stroke Scale [NIHSS] score,13 0-3) or high-risk TIA (ABCD2 score,14 ≥4). Brain imaging was required before randomization to rule out intracranial hemorrhage or a nonvascular etiology of the patient’s symptoms. Patients who were candidates for thrombolysis, thrombectomy, or carotid endarterectomy were excluded, as were patients requiring anticoagulant therapy. For this analysis, we excluded 598 patients enrolled outside the United States and 239 US patients with missing data on race or ethnicity.
The primary outcome of the POINT trial was the composite of ischemic stroke, myocardial infarction, or death from ischemic vascular causes. Ischemic stroke was a secondary trial end point. The primary and secondary efficacy outcomes of the POINT trial were adjudicated by an end point adjudication committee masked to the treatment assignments. Quiz Ref IDFor the analysis, we chose as our primary outcome postrandomization ischemic stroke, which was defined as: (1) the rapid onset of a new focal neurological deficit with clinical or imaging evidence of infarction and not attributable to a nonischemic etiology or (2) rapid worsening of an existing focal neurological deficit judged to be attributable to a new infarction.Quiz Ref ID Our secondary outcomes were lacunar vs nonlacunar stroke. Lacunar stroke was defined as a constellation of findings on the NIHSS consistent with a clinical lacunar syndrome: pure motor stroke, pure sensory stroke, mixed sensorimotor stroke, ataxic hemiparesis, or dysarthria/clumsy hand.15 The use of the NIHSS for distinguishing lacunar vs nonlacunar stroke has been previously validated.16
Our primary exposure variable was race as self-reported by trial participants. We classified the race of patients as black, white, or other and we separately adjusted for Hispanic ethnicity. The primary comparison of interest was for black vs white race. Other covariates were index event type (minor stroke vs TIA), study treatment assignment (clopidogrel vs placebo), other demographic characteristics (age and sex), traditional vascular risk factors (hypertension, diabetes, coronary artery disease, heart failure, valvular heart disease, and tobacco use), more detailed information on risk factors and stroke mechanisms (baseline systolic and diastolic blood pressure, baseline maximum carotid stenosis, baseline brain infarction on imaging results, and atrial fibrillation either at baseline or after randomization), medication use patterns (antihypertensive treatment before the index event and study drug adherence and statin treatment during follow-up), and the index event etiological subtype (lacunar vs nonlacunar). We included the NIHSS score as a covariate in a subgroup analysis of patients with an index minor ischemic stroke and the ABCD2 score as a covariate in a subgroup analysis of patients with an index TIA.
The etiological subtype of the index event was classified as lacunar if the patient’s presenting symptoms were consistent with a clinical lacunar syndrome.15 Carotid stenosis was recorded by investigators as the maximum detected on carotid ultrasonography, magnetic resonance angiography, computed tomographic angiography, or digital subtraction angiography and was fit into an ordinal variable categorized as (1) 1% to 49%, (2) 50% to 79%, (3) 80% to 99%, or (4) complete occlusion. Study drug adherence was defined based on pill counts. We ascertained incident atrial fibrillation based on adverse event reports and indications for study drug interruption (ie, patients who were switched to receiving anticoagulant therapy after the detection of incident atrial fibrillation).
We used a survival analysis to compare the risk of ischemic stroke among black and white individuals and patients of other races. Kaplan-Meier survival statistics and the log-rank test were used to compare the cumulative incidence of ischemic stroke among groups. A Cox proportional hazards analysis was used to examine the association between black race (compared with white race) with adjustment for covariates. We used stepwise models, first adjusting only for Hispanic ethnicity and then progressively adding index event type (minor stroke vs TIA) and study treatment assignment (aspirin plus clopidogrel vs aspirin alone); age and sex; traditional vascular risk factors (hypertension, diabetes, coronary artery disease, heart failure, valvular heart disease, and tobacco use); systolic and diastolic blood pressure, carotid stenosis, brain infarction, and a time-varying covariate for atrial fibrillation; medication use patterns; and the index event etiological subtype. We used the most fully adjusted model to perform subgroup analyses that were stratified by index event type (minor ischemic stroke vs TIA), with additional adjustment for the NIHSS stroke scale for patients with an index minor ischemic stroke or the ABCD2 score for patients with an index TIA. Given prior findings of an age-race interaction in black-white racial disparities in stroke incidence,1 we checked for an interaction between age and race. We also checked for interactions between study treatment assignment (clopidogrel vs placebo) and race, lacunar vs nonlacunar index event type and race, and baseline systolic blood pressure and race. All statistical analyses were performed using Stata/MP (version 15; StataCorp). The threshold of statistical significance was set at α = .05.
First, we performed an analysis in which we excluded patients from sites that did not enroll at least 1 black trial participant. Second, we performed an analysis in which we created a propensity score to predict black vs nonblack race based on all the baseline covariates described previously and adjusted a Cox model for this propensity score. Third, we included all patients regardless of missing race/ethnicity data and applied multiple imputation using chained equations17 to impute missing race/ethnicity variables based on all the baseline covariates described previously.
Of the 4881 POINT trial participants, we excluded 598 patients who were enrolled outside the United States and 239 US patients who had missing race/ethnicity data. The 239 excluded US patients had broadly similar baseline characteristics compared with the rest of the US patients (eTable 1 in the Supplement). Among the 4044 patients included in the analysis, 2942 (72.7%) were white, 918 (22.7%) were black, and 184 (4.5%) were of other races. Compared with white patients, black patients were younger, more often women, less often Hispanic; had a higher prevalence of hypertension and diabetes; more often were treated with antihypertensive agents before their index event; had a lower prevalence of tobacco use; had a lower degree of carotid stenosis; and more often had a lacunar syndrome as their index event (Table 1).
During the 90-day follow-up period, 232 ischemic strokes occurred in the 4044 patients. The cumulative risk of stroke was significantly higher among black patients (8.5%; 95% CI, 6.9%-10.6%) than among white patients (5.3%; 95% CI, 4.5%-6.2%) (P = .002 by the log-rank test; Figure). In a Cox model adjusted for Hispanic ethnicity, black race was associated with a significantly higher hazard of ischemic stroke compared with white race (hazard ratio [HR], 1.6; 95% CI, 1.2-2.2), whereas no increased risk was seen for other races (HR, 1.0; 95% CI, 0.5-1.9).
After stepwise adjustment for covariates, no notable change was seen in the association between black race and ischemic stroke (Table 2). In the final model, which included adjustment for index event type, study treatment assignment, demographics, risk factors, medication use patterns, and index event etiological subtype, black race was still associated with a significantly higher risk of ischemic stroke (HR, 1.6; 95% CI, 1.1-2.3) (Table 2). The association between black race and a higher risk of recurrence was broadly similar when examining secondary outcomes of lacunar stroke (HR, 2.1; 95% CI, 0.9-4.8) vs nonlacunar stroke (HR, 1.6; 95% CI, 1.1-2.4).
The associations were similar in subgroup analyses of patients with an index TIA (HR, 1.6; 95% CI, 0.98-2.8) vs an index minor ischemic stroke (HR, 1.5; 95% CI, 1.1-2.2). We also found no evidence of a significant interaction between age and black race, between study treatment assignment and black race, between lacunar vs nonlacunar index event and black race, or between baseline systolic blood pressure and black race in regard to the association with ischemic stroke (eTable 2 in the Supplement). Our findings were similar in sensitivity analyses limited to sites that enrolled black patients (HR, 1.6; 95% CI, 1.1-2.3), using a propensity score adjustment (HR, 1.5; 95% CI, 1.1-2.1), and imputing missing race/ethnicity (HR, 1.6; 95% CI, 1.2-2.1).
Quiz Ref IDAmong US participants in the POINT trial, we found that black patients faced a significantly higher risk of early ischemic stroke recurrence after a minor ischemic stroke or TIA. This difference applied to lacunar and nonlacunar outcome events and was not attenuated by adjustment for other demographic characteristics, risk factors, medication adherence, or index event etiological subtype. Black race was significantly associated with higher recurrence risk even in analyses limited to sites that enrolled black patients.
Prior studies have reported conflicting findings about whether black individuals in the United States experience a higher risk of recurrent stroke than white individuals. Analyses of 2 randomized clinical trials with populations enrolled in the subacute phase after an index ischemic stroke found no association between black race and recurrent stroke risk.8,9 In a large longitudinal cohort study, black race was also not found to be a risk factor for ischemic stroke recurrence, although early recurrences were not captured.6 However, 2 large, population-based studies did find an association between black race and stroke recurrence.5,7 These 2 studies used administrative claims data and thus may have missed early in-hospital recurrences and misclassified later nonstroke hospitalizations as recurrent strokes.5 Administrative claims data also do not allow granular adjustment for covariates such as blood pressure, medication adherence, or etiological subtype. In this context, our study provides novel findings that US black adults face a higher risk of early ischemic stroke recurrence compared with white individuals and that this disparity does not seem to be explained by established stroke risk factors, medication adherence, or lacunar vs nonlacunar etiological subtype.
Quiz Ref IDOur study has limitations. First, our study involved a population enrolled in a clinical trial and thus may not be generalizable to the rest of the US population. Patients enrolled in clinical trials tend to have a more favorable baseline prognosis than nonenrolled patients and may experience a positive participation effect18; POINT patients were strongly encouraged to adhere to standard secondary prevention practices, which are poorly adhered to in routine clinical practice,19 so our findings may be a conservative estimate of the true black-white racial disparity in recurrent stroke in the United States. Second, we lacked follow-up data beyond 90 days, and longer-term follow-up may have been illuminating. Third, our results may be affected by unmeasured factors, such as socioeconomic status, prior stroke or preexisting disability, the extent of cerebral white matter disease, or the severity or treatment of patients’ vascular risk factors, such as hypertension, for which we lacked data during follow-up. We were thus unable to conclusively delineate the reasons for the observed black-white disparities in stroke recurrence.
Our findings have several potential explanations. Black patients might receive less optimal care in the short-term setting after a minor ischemic stroke or TIA. Although the study population was closely monitored in the setting of a clinical trial, the POINT trial investigators may not have been directly managing risk factors, such as blood pressure, and such crucial facets of secondary stroke prevention may have differed systematically between black and white individuals. Furthermore, for any given degree of blood pressure elevation, black individuals may face a higher risk of stroke than white individuals,20 and this may be one mechanism by which black patients in POINT faced a higher recurrence risk even after adjustment for this important risk factor. However, the similar magnitude of risk even in normotensive patients suggests that the higher risk of recurrence in black individuals cannot be attributed only to differences in blood pressure. In addition, black patients in our sample may have had a higher prevalence of other high-risk stroke mechanisms compared with white individuals. The black-white disparity in stroke recurrence in POINT did not seem to be explained by lacunar pathophysiology, extracranial carotid atherosclerosis, or clinically apparent atrial fibrillation, but we were unable to assess other high-risk mechanisms, such as intracranial atherosclerosis or atrial cardiopathy, that may be more common in black patients.21,22 Our findings justify further research into racial differences in the underlying mechanisms of recurrent stroke. In the meantime, our findings imply that extra effort should be made to ensure that black patients with TIA or minor ischemic stroke have access to proven secondary prevention measures.11
Accepted for Publication: December 18, 2019.
Corresponding Author: Hooman Kamel, MD, Clinical and Translational Neuroscience Unit, Feil Family Brain and Mind Research Institute and Department of Neurology, Weill Cornell Medicine, 420 E 70th St, LH-413, New York, NY 10021 (firstname.lastname@example.org).
Published Online: February 24, 2020. doi:10.1001/jamaneurol.2020.0010
Author Contributions: Dr Kamel had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Kamel, Kleindorfer, Johnston.
Acquisition, analysis, or interpretation of data: Kamel, Zhang, Levitan, V. Howard, G. Howard, Soliman.
Drafting of the manuscript: Kamel.
Critical revision of the manuscript for important intellectual content: Zhang, Kleindorfer, Levitan, V. Howard, G. Howard, Soliman, Johnston.
Statistical analysis: Kamel, Zhang, Levitan.
Obtained funding: Kamel, Johnston.
Administrative, technical, or material support: Kamel, Johnston.
Conflict of Interest Disclosures: Dr Kamel reported serving as coprincipal investigator for the National Institutes of Health (NIH)–funded ARCADIA trial, which receives an in-kind study drug from the BMS-Pfizer Alliance and in-kind study assays from Roche Diagnostics, a steering committee member of Medtronic's Stroke AF trial (uncompensated), an end point adjudication committee member for a trial of empagliflozin for Boehringer-Ingelheim, and an advisory board member for Roivant Sciences on the topic of Factor XI inhibition. Dr Levitan reported grants from Amgen and personal fees from Novartis outside the submitted work. Dr Johnston reported nonfinancial support from Sanofi and grants from the NIH/National Institute of Neurological Disorders and Stroke (NINDS) and AstraZeneca. No other disclosures were reported.
Funding/Support: The POINT trial was funded by NIH/NINDS (grants U01NS062835, U01NS056975, and U01NS059041). Dr Kamel is supported by the NIH (grants R01NS097443 and R01HL144541).
Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: Dr Kamel is a Deputy Editor of JAMA Neurology, but he was not involved in any of the decisions regarding review of the manuscript or its acceptance.