Does Socioeconomic Status or Acculturation Modify the Association Between Ethnicity and Hypertension Treatment Before Stroke? Deborah A. Levine, MD, MPH; Lewis B. Morgenstern, MD; Kenneth M. Langa, MD, PhD; Lesli E. Skolarus, MD, MS; Melinda A. Smith, DrPH; Lynda D. Lisabeth, PhD Downloaded from http://stroke.ahajournals.org/ by guest on April 10, 2018 Background and Purpose Socioeconomic status and acculturation may modify the association between ethnicity and hypertension treatment before stroke. We assessed prestroke treatment of hypertension by ethnicity, education (proxy for socioeconomic status), and English proficiency (EP; proxy for acculturation) in a population-based stroke surveillance project. Methods Among 763 patients with first-ever stroke aged 45 years in the Brain Attack Surveillance in Corpus Christi project from 2000 to 2006, we examined self-reported hypertension treatment at the time of the stroke by ethnicity ( American [MA] versus non Hispanic white [NHW]) in the overall sample, within education strata (<high school, high school, >high school), and after dichotomizing MAs by self-reported EP (limited versus proficient). Logistic regression adjusted associations for age, sex, education, diabetes mellitus, coronary artery disease, hypercholesterolemia, and health insurance. Results NHWs and MAs reported similar hypertension treatment (84% versus 86%; P=0.53). Hypertension treatment was 84% for NHWs and 90% for MAs (P=0.18) in <high school stratum, 87% for NHWs and 75% for MAs (P=0.07) in high school stratum, and 81% for NHWs and 78% for MAs (P=0.73) in >high school stratum (ethnicity-by-education interaction, P=0.09). Hypertension treatment was 83% for NHWs, 87% for MAs with EP (P vsnhws =0.35), and 90% for MAs with limited EP (P vsnhws =0.13; ethnicity-by-ep interaction, P=0.22). Hypertension treatment was lower in uninsured patients (adjusted odds ratio, 0.13; 95% confidence interval, 0.03-0.60) or those with no physician visit 6 months (adjusted odds ratio, 0.09; 95% confidence interval, 0.03-0.24). Conclusions We found no evidence that socioeconomic status or acculturation modifies the association between ethnicity and hypertension treatment before stroke. (Stroke. 2013;44:3243-3245.) Key Words: acculturation ethnicity hypertension socioeconomic status treatment Ethnic disparities in the treatment of stroke risk factors may be difficult to detect 1 3 and may only become evident when examining those with low socioeconomic status (SES) or low levels of acculturation. 4 (MAs) comprise 66% of all the US Hispanics and are more likely to have low SES or less acculturation. 4 It is unknown whether low SES or less acculturation exacerbates MA-white differences in hypertension treatment before stroke. We examined prestroke treatment of hypertension stratified by ethnicity, education (a valid marker of SES), 5 and English proficiency (EP; a valid proxy for acculturation) 6 in a population-based, biethnic stroke surveillance project. Methods Study Population Stroke cases presenting between January 2000 and June 2006 were ascertained in Nueces County, TX, in the Brain Attack Surveillance in Corpus Christi project. 7 Through active and passive surveillance, Brain Attack Surveillance in Corpus Christi ascertains all cases of acute cerebrovascular disease presenting to the emergency department or directly admitted to any of the 7 hospitals in Nueces County. During the study period, there was an out-of-hospital surveillance program. 8 Out-of-hospital events accounted for a small percentage of acute strokes and were similar by ethnicity. 3 Trained abstractors verify stroke diagnoses on the basis of rigorous criteria. Neurologists validate stroke cases using source documentation and international clinical criteria. 9 At the time of their stroke hospitalization, patients or proxies for patients unable to participate completed an in-person, structured interview. Bilingual abstractors conducted the interview Received July 30, 2013; accepted August 2, 2013. From the Division of General Medicine and HSR&D Center for Clinical Management Research, Veterans Affairs Ann Arbor Healthcare System, Ann Arbor, MI (D.A.L., K.M.L.); and Stroke Program (D.A.L., L.B.M., L.E.S., M.A.S., L.D.L.), Department of Epidemiology (L.B.M., L.D.L.), Institute for Social Research (K.M.L.), University of Michigan, Ann Arbor, MI. The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/strokeaha. 113.003051/-/DC1. Correspondence to Deborah A. Levine, MD, MPH, University of Michigan, NCRC Bldg 16 Rm 430W, 2800 Plymouth Rd, Ann Arbor, MI 48109-2800. E-mail deblevin@umich.edu 2013 American Heart Association, Inc. Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.113.003051 3243
3244 Stroke November 2013 Downloaded from http://stroke.ahajournals.org/ by guest on April 10, 2018 in English or Spanish per patient preferences. Interview participation was similar by ethnicity. 7 Patients with their first Brain Attack Surveillance in Corpus Christi stroke (ischemic or hemorrhagic) were included. This project was approved by the University of Michigan and both Corpus Christi Health Systems Institutional Review Boards. Informed consent was obtained from all subjects. Variables Patients reported prestroke hypertension (physician diagnosis of high blood pressure ever) and its treatment (use of antihypertensive medication) at the time of the index stroke. All covariates, including race-ethnicity (non Hispanic white [NHW], MA), were self-reported except insurance status (medical record) and selected using the Andersen Behavioral Model framework. 10 EP was speaking English only or bilingual (more accultured). Limited EP was speaking Spanish only (less accultured). Months from last physician visit to stroke (0 6, 6 12, 12 24, >24, and never) were recorded. Of 1111 stroke cases identified, we excluded 328 with previous stroke/transient ischemic attack, 89 with other race-ethnicity, and 26 with no information on the outcome, leaving 763 cases in the study. Statistical Analysis We evaluated correlates of ethnicity (NHWs=referent) and hypertension treatment status using χ 2 test or ANOVA. We used multivariable logistic regression models to examine the association between ethnicity and hypertension treatment before and after adjusting for covariates: (1) overall and within education strata (<high school, high school, high school) and (2) after stratifying MAs by EP. We also compared hypertension treatment by ethnicity among those with diabetes mellitus. 11 Income was significantly correlated with education (coefficient 0.54; P<0.001) so we included education alone and did sensitivity analyses adjusting for income. 4 We examined interaction terms (ethnicity-by-education and ethnicity-by-ep) in the adjusted models. Table. Odds Ratios (95% Confidence Intervals) for Hypertension Treatment, Unadjusted and Adjusted: BASIC, 2000 to 2006 Unadjusted Results Hypertension treatment was similar in MAs and NHWs overall (86% versus 84%; P=0.53) and in those with diabetes mellitus (91% versus 89%; P=0.70). Online-only Data Supplement presents patient characteristics by ethnicity and hypertension treatment status. Hypertension treatment was 84% for NHWs and 90% for MAs (P=0.18) in <high school stratum, 87% for NHWs and 75% for MAs (P=0.07) in high school stratum, and 81% for NHWs and 78% for MAs (P=0.73) in >high school stratum (ethnicity-by-education interaction, P=0.09). Hypertension treatment was 83% for NHWs, 87% for MAs with EP (P vsnhws =0.35), and 90% for MAs with limited EP (P vsnhws =0.13; ethnicity-by-ep interaction, P=0.22). MAs had similar likelihood of hypertension treatment before stroke overall, within education strata, and after dichotomizing MAs by EP (Table; online-only Data Supplement). Hypertension treatment was lower in uninsured patients (adjusted odds ratio, 0.13; 95% confidence interval, 0.03 0.60) or those with no physician visit 6 months (adjusted odds ratio, 0.09; 95% confidence interval, 0.03 0.24). There were no ethnic differences in treatment in analyses that used medical record diagnosis of hypertension or that included income (online-only Data Supplement). Discussion We found no evidence that SES or acculturation modifies the association between ethnicity and hypertension treatment before stroke. Partially Adjusted: Includes Sociodemographics, Comorbidity, and Health Insurance Fully Adjusted: Includes Healthcare Use* Overall sample n 520 416 289 American 1.17 (0.72 1.88) 0.93 (0.49 1.78) 1.33 (0.60 2.97) c-statistic 0.52 0.70 0.78 Less than high school education n 245 191 141 American 1.76 (0.77 4.06) 1.85 (0.70 4.90) 2.98 (0.85 10.46) c-statistic 0.56 0.74 0.83 High school education n 149 118 81 American 0.46 (0.20 1.08) 0.45 (0.15 1.42) 1.24 (0.26 5.94) c-statistic 0.59 0.64 0.77 More than high school education n 126 107 67 American 0.83 (0.30 2.34) 0.82 (0.18 3.85) 0.71 (0.06 8.72) c-statistic 0.52 0.85 0.92 Partially adjusted models included age, sex, education, coronary artery disease, diabetes mellitus, hypercholesterolemia, and no health insurance. Fully adjusted models also included no primary care physician and last physician visit. BASIC indicates Brain Attack Surveillance in Corpus Christi. *Available from January 2000 to 2004 only.
Levine et al Ethnicity and Hypertension Treatment Before Stroke 3245 Downloaded from http://stroke.ahajournals.org/ by guest on April 10, 2018 Several factors potentially explain our findings. Most MAs in our study were born in the United States and insured (93%). Health insurance accounts for the largest proportion of the Hispanic-white disparity in access to physician care. 12 Health insurance and access to care differences may be greater in communities with more MA immigrants. Alternatively, we may have been unable to detect small ethnic differences in prestroke treatment because of sample size (n=520). On the basis of a post hoc analysis, we had 80% power (α=0.05) to detect an absolute hypertension treatment rate for MAs that was either 10% lower or 8% higher than that of NHWs (84%). Finally, ethnic differences in treatment of vascular risk factors at the population-level may vary by geographic region and may even be absent in some communities or subpopulations. 13 Our study has limitations. We lacked information on control or duration of hypertension. Hispanics have worse hypertension control than whites, and this gap persisted between 1999 and 2006. 14 Despite similar treatment rates, MAs may have worse prestroke hypertension control than NHWs. Moreover, our study does not preclude that ethnic differences in hypertension prevalence, treatment, or control contribute to MAs higher stroke incidence. 3 Neither SES nor acculturation amplified ethnic differences in hypertension treatment. There was a tendency for MAs and NHWs with higher SES to have lower hypertension treatment rates before stroke, which is counterintuitive to expected trends. Sources of Funding National Institutes of Health R01NS38916 supported this work. Disclosures National Institutes of Health funding also included K23AG040278 and P30DK092926 (Dr Levine), R01NS38916 (Drs Morgenstern, Lisabeth, and Smith), and K23NS073685 (Dr Skolarus). The other authors report no conflicts. References 1. Lisabeth LD, Smith MA, Sánchez BN, Brown DL. Ethnic disparities in stroke and hypertension among women: the BASIC project. Am J Hypertens. 2008;21:778 783. 2. Smith MA, Risser JM, Lisabeth LD, Moyé LA, Morgenstern LB. Access to care, acculturation, and risk factors for stroke in : the Brain Attack Surveillance in Corpus Christi (BASIC) project. Stroke. 2003;34:2671 2675. 3. Morgenstern LB, Smith MA, Lisabeth LD, Risser JM, Uchino K, Garcia N, et al. Excess stroke in compared with non-hispanic Whites: the Brain Attack Surveillance in Corpus Christi Project. Am J Epidemiol. 2004;160:376 383. 4. Levine DA, Allison JJ, Cherrington A, Richman J, Scarinci IC, Houston TK. Disparities in self-monitoring of blood glucose among low-income ethnic minority populations with diabetes, United States. Ethn Dis. 2009;19:97 103. 5. Winkleby MA, Jatulis DE, Frank E, Fortmann SP. Socioeconomic status and health: how education, income, and occupation contribute to risk factors for cardiovascular disease. Am J Public Health. 1992;82:816 820. 6. Coronado GD, Thompson B, McLerran D, Schwartz SM, Koepsell TD. A short acculturation scale for -American populations. Ethn Dis. 2005;15:53 62. 7. Smith MA, Risser JM, Moyé LA, Garcia N, Akiwumi O, Uchino K, et al. Designing multi-ethnic stroke studies: the Brain Attack Surveillance in Corpus Christi (BASIC) project. Ethn Dis. 2004;14:520 526. 8. Morgenstern LB Smith MA, Sánchez BN, Brown DL, Zahuranec DB, Garcia N, et al. Persistent ischemic stroke disparities despite declining incidence in. Ann Neurol. August 13, 2013. doi:10.1002/ana.23972. 9. Asplund K, Tuomilehto J, Stegmayr B, Wester PO, Tunstall-Pedoe H. Diagnostic criteria and quality control of the registration of stroke events in the MONICA project. Acta Med Scand Suppl. 1988;728:26 39. 10. Gelberg L, Andersen RM, Leake BD. The Behavioral Model for Vulnerable Populations: application to medical care use and outcomes for homeless people. Health Serv Res. 2000;34:1273 1302. 11. Chen G, McAlister FA, Walker RL, Hemmelgarn BR, Campbell NR. Cardiovascular outcomes in Framingham participants with diabetes: the importance of blood pressure. Hypertension. 2011;57:891 897. 12. Mahmoudi E, Jensen GA. Diverging racial and ethnic disparities in access to physician care: comparing 2000 and 2007. Med Care. 2012;50:327 334. 13. Bonds DE, Zaccaro DJ, Karter AJ, Selby JV, Saad M, Goff DC Jr. Ethnic and racial differences in diabetes care: The Insulin Resistance Atherosclerosis Study. Diabetes Care. 2003;26:1040 1046. 14. McWilliams JM, Meara E, Zaslavsky AM, Ayanian JZ. Differences in control of cardiovascular disease and diabetes by race, ethnicity, and education: U.S. trends from 1999 to 2006 and effects of medicare coverage. Ann Intern Med. 2009;150:505 515.
Does Socioeconomic Status or Acculturation Modify the Association Between Ethnicity and Hypertension Treatment Before Stroke? Deborah A. Levine, Lewis B. Morgenstern, Kenneth M. Langa, Lesli E. Skolarus, Melinda A. Smith and Lynda D. Lisabeth Downloaded from http://stroke.ahajournals.org/ by guest on April 10, 2018 Stroke. 2013;44:3243-3245; originally published online September 17, 2013; doi: 10.1161/STROKEAHA.113.003051 Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright 2013 American Heart Association, Inc. All rights reserved. Print ISSN: 0039-2499. Online ISSN: 1524-4628 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://stroke.ahajournals.org/content/44/11/3243 Data Supplement (unedited) at: http://stroke.ahajournals.org/content/suppl/2013/09/17/strokeaha.113.003051.dc1 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Stroke is online at: http://stroke.ahajournals.org//subscriptions/
SUPPLEMENTAL MATERIAL Does SES or Acculturation Modify the Association between Ethnicity and Hypertension Treatment before Stroke? Deborah A. Levine MD, MPH 123 ; Lewis B. Morgenstern MD 34 ; Kenneth M. Langa MD, PhD 125 ; Lesli E. Skolarus, MD, MS 3 ; Melinda A. Smith, DrPH 3 ; Lynda D. Lisabeth, PhD 34 1 Division of General Medicine, University of Michigan Health System, and Ann Arbor VA Healthcare System, Ann Arbor, MI 2 Veterans Affairs Health Services Research and Development Center of Excellence, Ann Arbor, MI 3 Stroke Program, University of Michigan Medical School, Ann Arbor, MI 4 Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, MI 5 Institute for Social Research, University of Michigan, Ann Arbor, MI
Supplemental Table I: Characteristics and Outcome by Ethnicity (N=763): BASIC, 2000-2006 Characteristic Non-Hispanic Whites n=376 (49) n=387 (51) P-value Age, mean ± SD, years 73.6 ± 12.0 67.0 ± 12.2 <0.001 Age 45-64 years 82 (22) 167 (43) <0.001 Women 201 (53) 203 (52) 0.78 No health insurance 8 (2) 28 (7) <0.001 Language proficiency <0.001 English proficiency 310 (99) 234 (70) Limited English proficiency 2 (1) 100 (30) Education <0.001 Less than high school 87 (23) 261 (67) High school 140 (37) 83 (21) More than high school 149 (40) 43 (11) Income <0.001 Less than $10,000 61 (16) 160 (41) $10,000-$19,999 61 (16) 107 (28) $20,000-$29,999 63 (17) 36 (9) $30,000-$49,000 53 (14) 33 (9) $50,000 or more 63 (17) 12 (3) Missing/unreported 75 (20) 39 (10) Proxy respondent 133 (35) 125 (32) 0.37 Clinical factors Coronary heart disease 108 (29) 107 (28) 0.67 Diabetes 92 (27) 183 (54) <0.001 High cholesterol 155 (50) 145 (50) 1.00 Access to Care (n=595)** No primary care doctor 24 (9) 42 (14) 0.04 Last doctor visit >6 months or 34 (13) 40 (14) 0.68 never Last doctor visit 6 months 233 (87) 247 (86) Outcomes Self-reported hypertension 253 (67) 267 (69) 0.61 Use of anti-hypertensive medication among those with self-reported hypertension 212 (84) 229 (86) 0.53 *Language proficiency was not asked from July 1, 2005 to May 26, 2006. **Available 2000-2004 only.
Supplemental Table II: Proportions and Unadjusted Odds Ratios (95% Confidence Intervals) of Pre-Stroke Hypertension Treatment by Subgroups of Stroke Patients with Hypertension (N=520): BASIC, 2000-2006 Untreated Hypertension n=79 (15) Treated Hypertension n=441 (85) Unadjusted Odds Ratio (95% CI) Subgroup P-value American 38 (14) 229 (86) 0.53 1.17 (0.72-1.88) Non-Hispanic White 41 (16) 212 (84) Referent Age, mean ± SD, years 66.5 ± 12.2 71.2 ± 12.0 0.002 NA Age 45-64 years 35 (21) 129 (79) 0.008 0.52 (0.32-0.85) Age 65 years 44 (12) 312 (88) Referent Women 37 (13) 248 (87) 0.12 1.46 (0.90-2.36) Men 42 (18) 193 (82) Referent No health insurance 8 (38) 13 (62) 0.003 0.27 (0.11-0.67) Health insurance 70 (14) 427 (86) Referent English proficiency 55 (15) 302 (85) 0.29 0.66 (0.30-1.44) Limited English proficiency 8 (11) 67 (89) Referent Education 0.07 < high school 28 (11) 217 (89) Referent High school 26 (17) 123 (83) 0.61 (0.34-1.09) >high school 25 (20) 101 (80) 0.52 (0.29-0.94) Income 0.16 <$10,000 17 (11) 134 (89) Referent $10,000-$19,999 21 (17) 200 (83) 0.60 (0.30-1.20) $20,000-$29,999 10 (16) 51 (84) 0.65 (0.28-1.51) $30,000-$49,000 14 (25) 43 (75) 0.39 (0.18-0.86) $50,000 or more 9 (18) 42 (82) 0.59 (0.25-1.43) Missing/unreported 8 (10) 71 (90) 1.13 (0.46-2.74) Proxy-respondent 13 (8) 156 (92) <0.001 2.80 (1.50-5.24) Self-respondent 66 (19) 283 (81) Referent Clinical factors Current cigarette smoking 20 (23) 67 (77) 0.03 0.54 (0.30-0.95) No current cigarette smoking 58 (14) 363 (86) Referent Coronary heart disease 14 (8) 157 (92) 0.002 2.56 (1.39-4.72) No coronary heart disease 64 (19) 280 (81) Referent Diabetes 21 (10) 195 (90) 0.008 2.08 (1.20-3.59) No diabetes 49 (18) 219 (82) Referent Hypercholesterolemia 32 (13) 217 (87) 0.04 1.70 (1.01-2.83) No hypercholesterolemia 38 (20) 152 (80) Referent Access to Care (n=595) No primary care doctor 11 (42) 15 (58) <0.001 0.20 (0.09-0.47) Primary care doctor 46 (13) 309 (87) Referent Last doctor visit >6 months or 23 (61) 15 (39) <0.001 0.07 (0.04-0.15) never
Last doctor visit 6 months 34 (10) 306 (90) Referent These are row percentages. *Language proficiency was not asked from July 1, 2005 to May 26, 2006. **Available 2000-2004 only.
Supplemental Table III: Hypertension Treatment by Ethnicity and Education: BASIC, 2000-2006 Less Than High School Education Non-Hispanic Whites n=87 (25) n=261 (75) P-value Outcome Self-reported hypertension 62/87 (71) 183/261 (70) 0.84 Use of anti-hypertensive medication among those with self-reported hypertension High School Education Outcome 52/62 (84) 165/183 (90) 0.18 Non-Hispanic Whites n=140 (63) n=83 (37) P-value Self-reported hypertension 92/140 (66) 57/83 (69) 0.65 Use of anti-hypertensive medication among those with self-reported hypertension More Than High School Education 80/92 (87) 43/57 (75) 0.07 Non-Hispanic Whites n=149 (78) n=43 (22) P-value Outcome Self-reported hypertension 99/149 (66) 27/43 (63) 0.66 Use of anti-hypertensive medication among those with self-reported hypertension 80/99 (81) 21/27 (78) 0.73
Supplemental Table IV: Hypertension Treatment by Ethnicity and English Proficiency: BASIC, 2000-2006 Non- Hispanic Whites n=310 (48) with English Proficiency n=234 (36) P-value vs. Non- Hispanic Whites with limited English Proficiency n=100 (15) P-value vs. Non-Hispanic Whites Self-reported 200/310 (65) 157/234 (67) 0.53 73/100 (73) 0.12 hypertension Use of antihypertensive medication among those with self-reported hypertension 166/200 (83) 136/157 (87) 0.35 66/73 (90) 0.13
Supplemental Table V: Self-Reported Anti-Hypertensive Medication Treatment among Stroke Patients with Medical Record Diagnosis of Hypertension by Ethnicity: BASIC, 2000-2006 Characteristic Medical record diagnosis of hypertension Use of anti-hypertensive medication among those with medical record diagnosis of hypertension Non-Hispanic Whites n=376 (49) n=387 (51) P-value 270 (72) 282 (73) 0.74 201/227 (89) 217/247 (88) 0.82
VI. Additional Statistical Analyses Hypertension Treatment among those with Diabetes Mellitus Given that hypertension is the major predictor of vascular events in adults with diabetes 27 and diabetes is more common in than non-hispanic whites, we also compared hypertension prevalence and treatment by ethnicity among stroke patients with diabetes. Among those with diabetes, hypertension treatment was 91% in and 89% in non-hispanic whites (P=0.70). Temporal Trends in Hypertension Treatment For each ethnic group, the frequency of the outcome measure across the individual years was compared by using Cochran-Mantel-Haenszel χ 2 test for trend in the overall cohort, after stratification by education, and also after dichotomization of by EP. The treatment of pre-stroke hypertension was stable from 2000 to 2006 (P from Mantel- Haenszel Chi-Square=0.38). There was no evidence of an ethnicity-by-time interaction when comparing non-hispanic whites to in the overall cohort (P=0.07) or after dichotomizing the latter group by EP (P 0.41). Small numbers precluded examination of ethnicity-by-time interactions within education strata. There was no evidence of an education-bytime interaction (P=0.59).