Abstract
Introduction
African Americans experience greater post-stroke disability than whites. We explored whether these differences are due to differences in pre-stroke function.
Methods
The Panel Study of Income Dynamics (PSID) is a nationally representative US panel survey of families and their descendants. We included all PSID respondents who reported an incident stroke between 2001 and 2011. Our primary outcome was an index representing the sum of total activities of daily living (ADL) limitations (0–7) and the secondary outcome was an index of instrumental activities of daily living (IADL) limitations (0–6). Survey-weighted descriptive statistics and Poisson regression were used to estimate racial differences in ADL and IADL before, with and after the wave when incident stroke was reported.
Results
A total of 534 incident strokes were identified, 198 (37%) in African-Americans. There were no pre-stroke racial differences in activity limitations (0.7 vs. 0.7, p=0.99). In the wave of the incident stroke (between 0–2 years from incident stroke), African Americans had considerably more ADL limitations than whites (2.2 vs. 1.5, p=0.048). These racial differences persisted after adjusting for age, sex and comorbidities. For IADLs, adjusted models suggested small pre-stroke racial differences and larger post-stroke differences.
Conclusion
Racial disparities in post-stroke ADL limitations are not due to pre-stroke activity limitations. Instead, differences appear largest in the first 2 years after stroke
Introduction
African-American stroke survivors have worse functional and cognitive outcomes than whites and more activity limitations.1,2 While these racial differences are largely attributable to lower post-stroke physical capacity,2 the reasons for capacity differences are unknown. Factors such as severity,3,4 underlying mechanism,5 and treatments are similar by race and would not account for the large differences in functioning.6
An alternative explanation is that African-Americans have lower pre-stroke function than whites. One study found that before their stroke older African-American stroke survivors had more limitations in instrumental activities of daily living (IADL) compared to whites.7 Given that African-Americans have strokes at considerably younger ages than whites, however, a question remains as to whether such findings are robust when all ages are considered.8 We used a nationally representative sample of adult stroke survivors of all ages to examine racial differences in the prevalence of activity of daily living (ADL) limitations and IADLs preceding and following stroke.
Methods
This study draws upon the Panel Study of Income Dynamics (PSID), a prospective, nationally representative study of 5,000 US families in 1968 and their descendants. Interviews have been conducted annually through 1997 and biannually thereafter, typically with the household head or spouse. The PSID closely represents the US adult population with respect to health and socioeconomic status.9
Data on self-reported stroke were available from 1999–2011. Our sample included individuals who reported they were stroke-free in 1999. Incident stroke was identified by response to the question, “Has a doctor ever told {you/your spouse} that {you have/your spouse has} had a stroke?” The first wave in which a respondent answered affirmatively was assigned as the incident stroke wave.
Our primary outcome was an index variable representing the sum (0–7) of ADL limitations in each wave. Respondents received one point for each activity for which difficulty was reported (yes vs. no). Activities included bathing, dressing, toileting, eating, getting in/out of bed/chair, walking, and getting outside. A secondary outcome variable represented the sum (0–6) of IADL limitations (difficulty preparing meals, shopping, managing money, using the telephone, heavy and light housework due to a health problem). The primary exposure was self-reported race (African American vs. white).
Survey-weighted estimates of limitations were calculated by race for the wave prior to stroke, the incident wave, and up to 3 subsequent waves from stroke, and were repeated stratified by age (<50 vs. 50 and older) We also estimated associations between race and activity limitations for each year relative to the incident stroke, adjusting for age, sex and all PSID individual comorbidities (self-reported hypertension, diabetes, cancer, heart attack, arthritis, asthma, emotional or psychological diagnoses, lung disease) using Poisson regression. Racial differences were estimated from these models using average marginal effects.
Results
A total of 534 incident strokes were identified of which 198 (37%) were among African Americans. The population is summarized in Table 1. There were no racial differences in the proportion of stroke survivors who died (10.1% in African-Americans vs. 8.3 in whites, p = 0.49) or in the mean duration of follow-up (2.0 waves for African-Americans vs. 2.0 waves for whites, p=0.93).
Table 1.
Study Population
| White (n=336) | African- American (n=198) |
p | |
|---|---|---|---|
| N (%) | N (%) | ||
| Demographics | |||
| Age mean(SD) | 60.1 (16.5) | 52.2 (14.5) | < 0.01 |
| Female | 178 (53.0%) | 122 (61.6%) | 0.05 |
| Comorbidities | |||
| Hypertension | 142 (42.3%) | 113 (57.1%) | < 0.01 |
| Diabetes | 48 (14.3%) | 46 (23.2%) | 0.01 |
| Cancer | 31 (9.2%) | 9 (4.5%) | 0.05 |
| MI | 41 (12.2%) | 13 (6.6%) | 0.04 |
| CAD | 67 (19.9%) | 23 (11.6%) | 0.01 |
| Arthritis | 27 (8.0%) | 24 (12.1%) | 0.12 |
| Asthma | 128 (38.1%) | 60 (30.3%) | 0.07 |
| Psych Diagnosis | 38 (11.3%) | 21 (10.6%) | 0.80 |
| Lung Disease | 27 (8.0%) | 12 (6.1%) | 0.40 |
| Pre-stroke Function | |||
| IADL Index mean(SD) | 0.4 (0.9) | 0.4 (1) | 0.71 |
| Activity Index mean(SD) | 0.7 (1.5) | 0.7 (1.4) | 0.99 |
There were no pre-stroke racial differences in ADL limitations. Racial differences arose in the incident stroke wave (between 0–2 years from incident stroke), African Americans had more ADL limitations than whites (2.2 vs. 1.5, p=0.048). This difference decreased over time. This pattern was slightly amplified after adjustment for controls. (Figure 1)
Figure 1.
ADL limitations by race and time. Panel A: unadjusted ADL limitations; Panel B: ADL limitations adjusted for age, sex and all comorbidities; Panel C: unadjusted ADL limitations stratified by age and race. * p < 0.05; ** p < 0.01.
In unadjusted analyses, the only racial difference in IADL limitations emerged two waves (4–6 years) after the incident stroke wave when African-Americans had more limitations than whites (1.6 vs. 0.8, p=0.04). After adjusting for age, sex and comorbidities, African Americans had marginally more pre-stroke IADL limitations (0.5 vs. 0.3, p= 0.03), and about twice as many IADL limitations in incident stroke wave (1.3 vs. 0.7, p < 0.01). These racial differences persisted (1.3 vs. 0.9, p=0.06 one wave and 1.7 vs. 0.8, p < 0.01 two waves after stroke). (Figure 2) In age-stratified analyses, there were no statistically significant racial differences in ADL or IADL limitations. However, among those ages 50 and older, patterns were consistent with small differences in pre-stroke IADL limitations, which widened at the time of incident stroke.
Figure 2.
IADL Limitations by race and time. Panel A: unadjusted IADL limitations; Panel B: IADL limitations adjusted for age, sex and all comorbidities; Panel C: unadjusted IADL limitations stratified by race and age category. * p < 0.05; ** p < 0.01.
Discussion
In this nationally representative study of over 500 stroke patients, there were no racial differences in pre-stroke ADL limitations. In contrast, marked racial differences in post-stroke ADL limitations were observed. Therefore it is unlikely that the observed racial differences in post-stroke ADL limitations are accounted for by differences in pre-stroke function.
We found that small racial differences in IADL limitations existed prior to stroke, widened at the time of the stroke and persisted. This finding differs from prior a study of individuals over age 50, which found pre-stroke racial differences in disability narrowed at the time of stroke. Among those over 50 we found that while small non-significant pre-stroke racial differences existed, these differences increased at the time of stroke. We conclude that differences in pre-stroke functioning may be less important for racial differences in post-stroke functioning than previously suggested.
The primary limitations of this study are that stroke was determined by self-report and the lack of stroke severity measures. Neither of these limitations should affect our primary conclusions as self-reported stroke has reasonable sensitivity and specificity10 and studies have suggested that stroke severity does not differ by race.3,4
Given that racial differences in post-stroke ADL and IADL limitations are largest within 2 years of stroke onset, efforts to reduce African-American activity limitations should focus on the time period immediately after stroke — in the hospital, in rehabilitation or upon return to the community. While there are no marked differences in stroke severity3,4,11,12 or measured processes of care by race,6 a number of alternate explanations may account for the early emergence of functional differences. For example, unmeasured differences in acute quality of care, access to care, differential utilization of post-acute rehabilitation or interaction with the local environment, possibly mediated by economic factors, are all credible and non-exclusive explanations for these differences. Understanding which, if any, of these explanations are most applicable is essential to improving stroke outcomes for African-Americans and may have broader implications for stroke survivorship.2
Acknowledgements
The data used in this study was partly collected with support by the National Institutes of Health (R01-AG040213, R01-HD069609) and the National Science Foundation (1157698). This work was also supported by NIH grants to Dr. Burke (K08 NS082597, R01 MD008879) and Dr. Skolarus (K23 NS073685, R01 MD008879).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Roth DL, Haley WE, Clay OJ, et al. Race and gender differences in 1-year outcomes for community-dwelling stroke survivors with family caregivers. Stroke. 2011 Mar;42(3):626–631. doi: 10.1161/STROKEAHA.110.595322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Burke JF, Freedman VA, Lisabeth LD, Brown DL, Haggins A, Skolarus LE. Racial differences in disability after stroke: results from a nationwide study. Neurology. 2014 Jul 29;83(5):390–397. doi: 10.1212/WNL.0000000000000640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kleindorfer D, Lindsell C, Alwell KA, et al. Patients living in impoverished areas have more severe ischemic strokes. Stroke. 2012 Aug;43(8):2055–2059. doi: 10.1161/STROKEAHA.111.649608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Qian F, Fonarow GC, Smith EE, et al. Racial and ethnic differences in outcomes in older patients with acute ischemic stroke. Circ Cardiovasc Qual Outcomes. 2013 May 1;6(3):284–292. doi: 10.1161/CIRCOUTCOMES.113.000211. [DOI] [PubMed] [Google Scholar]
- 5.White H, Boden-Albala B, Wang C, et al. Ischemic stroke subtype incidence among whites, blacks, and Hispanics: the Northern Manhattan Study. Circulation. 2005 Mar 15;111(10):1327–1331. doi: 10.1161/01.CIR.0000157736.19739.D0. [DOI] [PubMed] [Google Scholar]
- 6.Schwamm LH, Reeves MJ, Pan W, et al. Race/Ethnicity, Quality of Care, and Outcomes in Ischemic Stroke. Circulation. Am Heart Assoc. 2010 Apr 6;121(13):1492. doi: 10.1161/CIRCULATIONAHA.109.881490. [DOI] [PubMed] [Google Scholar]
- 7.Capistrant BD, Mejia NI, Liu SY, Wang Q, Glymour MM. The disability burden associated with stroke emerges before stroke onset and differentially affects blacks: results from the health and retirement study cohort. J Gerontol A Biol Sci Med Sci. 2014 Jul;69(7):860–870. doi: 10.1093/gerona/glt191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Howard VJ, Kleindorfer DO, Judd SE, et al. Disparities in stroke incidence contributing to disparities in stroke mortality. Ann Neurol. 2011 Apr;69(4):619–627. doi: 10.1002/ana.22385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Survey Research Center, Institute for Social Research. Panel Study of Income Dynamics, public use dataset [Internet] Ann Arbor MI: Available from: http://psidonline.isr.umich.edu/default.aspx. [Google Scholar]
- 10.Okura Y, Urban LH, Mahoney DW, Jacobsen SJ, Rodeheffer RJ. Agreement between self-report questionnaires and medical record data was substantial for diabetes, hypertension, myocardial infarction and stroke but not for heart failure. J Clin Epidemiol. 2004 Oct;57(10):1096–1103. doi: 10.1016/j.jclinepi.2004.04.005. [DOI] [PubMed] [Google Scholar]
- 11.Hartmann A, Rundek T, Mast H, Paik M, Boden B. Mortality and causes of death after first ischemic stroke: the Northern Manhattan Stroke Study. Neurology. 2001 doi: 10.1212/wnl.57.11.2000. [DOI] [PubMed] [Google Scholar]
- 12.Jones MR, Horner RD, Edwards LJ, et al. Racial variation in initial stroke severity. Stroke. 2000 Mar;31(3):563–567. doi: 10.1161/01.str.31.3.563. [DOI] [PubMed] [Google Scholar]