• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Health Care Poor Underserved. Author manuscript; available in PMC Mar 25, 2012.
Published in final edited form as:
PMCID: PMC3312013

Racial/ethnic Heterogeneity in the Socioeconomic Patterning of CVD Risk Factors: in the United States: The Multi-Ethnic Study of Atherosclerosis

Shawn Boykin, PhD, Research fellow, Ana V. Diez-Roux, MD, PhD, MPH, Professor and Director, Mercedes Carnethon, PhD, Associate professor, Sandi Shrager, MS, Analyst, Hanyu Ni, PhD, Research scientist, and Melicia Whitt-Glover, PhD, President


Many studies document racial variation, gender differences, and socioeconomic status (SES) patterning in cardiovascular disease (CVD) risk factors but few studies have investigated heterogeneity in SES differences by race/ethnicity or gender. Using data from the Multi-Ethnic Study of Atherosclerosis (N = 6,814) and stratified regression models, we investigated race/ethnic differences in the SES patterning of diabetes, hypertension, smoking, and body mass index (BMI). Inverse socioeconomic gradients in hypertension, diabetes, smoking, and BMI were observed in White and Black women but associations were weaker or absent in Hispanic and Chinese women (except in the case of diabetes for Hispanic women). Even greater heterogeneity in social patterning of risk factors was observed in men. In White men all four risk factors were inversely associated with socioeconomic position, although often associations were only present or were stronger for education than for income. The inverse socioeconomic patterning was much less consistent in men of other races/ethnic groups, and higher SES was associated with higher BMI in non-White men. These findings have implications for understanding the causes of social patterning, for the analysis of SES adjusted race/ethnic differences, and for the targeting of interventions.

Keywords: Cardiovascular disease, risk factors, socioeconomic status, race, ethnicity

Cardiovascular disease and its risk factors remain the most common causes of morbidity and mortality in minority populations and are a major cause of health disparities.1 Racial/ethnic variation in cardiovascular disease (CVD) risk has been widely documented in the United States (U.S.). This variation includes important differences in the prevalence, severity, and onset of risk factors by race/ethnicity. For example there is abundant evidence of higher prevalence of hypertension in Blacks than in other racial/ethnic groups2 and higher prevalence of diabetes in Hispanics than in other racial/ethnic groups.3 Body mass index and smoking are also strongly patterned by race/ethnicity.4 It is often noted that racial/ethnic differences in socioeconomic status contribute substantially to race differences in health.5 Supporting this assumption are abundant data showing higher levels of poverty, lower income, lower education, and lower occupational status in non-Whites than in Whites.5 In addition, racial/ethnic differences are usually reduced after statistical adjustment for measures of socioeconomic position, although some differences often remain.48

Williams et al.5 recently emphasized the need to move beyond adjusting for SES in studies of health disparities to examine specifically the complex ways in which SES combines with race/ethnicity to affect patterns of health. As they suggest, a layer of complexity is added when gender differences in health are simultaneously considered with race/ethnicity and SES. Despite this complexity, the vast majority of studies on the socioeconomic patterning of cardiovascular risk have focused on predominantly White samples.9 Few studies have had the racial/ethnic and gender heterogeneity necessary to examine specifically whether social inequalities in CVD risk are simultaneously modified by gender and race/ethnicity.

There are a number of reasons that the socioeconomic patterning of cardiovascular risk might vary by race/ethnicity or gender. Higher levels of SES could be differentially associated with environmental, psychosocial or behavioral correlates depending on other factors such as immigration history, place of residence, or exposure to discrimination. Indeed there is evidence that the socioeconomic patterning of cardiovascular disease has varied over time and across countries.10,11 Describing potential heterogeneity in SES patterning of CVD risk factors by race/ethnicity and gender is critical from a practical standpoint for appropriately targeting preventative interventions. More fundamentally, examining heterogeneity in SES patterning of CVD risk at the intersection of race and gender may be fundamental to understanding the social patterning of CVD risk generally and the fundamental causes of racial/ethnic disparities.

In addition to overall racial/ethnic differences in the SES patterning of cardiovascular risk, there may also be important differences related to the indicator of social position examined. For example, it is well established that economic returns on education in the U.S. are smaller for Blacks and Hispanics compared with Whites.12,13 Additionally, several studies have shown that among U.S. Blacks, higher education is associated with greater reports of discrimination and stress, both of which have been linked to worse health.14,1517 On the other hand, income and wealth may be less predictive of CVD risk factors for Blacks or Hispanics than for Whites since gains in income for these groups have not always been associated with improvements in access to care or access to other resources.15,18,19

This investigation takes advantage of a unique population-based multiethnic sample with detailed measures of socioeconomic position and cardiovascular risk to examine the socioeconomic patterning of cardiovascular risk in different racial/ethnic and gender groups, as well as the extent to which these differences vary depending on the socioeconomic indicator used. We hypothesized that SES gradients in the prevalence of CVD risk factors (that is, the systematic patterning from high to low education or income) would be strongest and most consistent in Whites, present and less consistent in Blacks (because of the added effects of factors such as discrimination), and absent in Hispanics and Chinese, reflecting the differential correlates of socioeconomic position in the countries of origin of largely immigrant groups. We also hypothesized that differences in social patterning by race/ethnicity would be more pronounced in the case of education than in the case of income because of previous work documenting that returns on education vary by race.12


Study population

Data were taken from the Multiethnic Study of Atherosclerosis (MESA), a population-based study of atherosclerosis in 6,814 men and women, aged 45–84 years at baseline in 2000–2002. Participants were selected from six sites including Manhattan and the Bronx, N.Y.; Baltimore, Md.; Forsyth County, N.C.; Chicago, Ill.; St. Paul, Minn.; and Los Angeles, Calif. People with a history of physician-diagnosed cardiovascular disease or prior CVD event were excluded. Approximately 38% of the sample was White, 28% Black, 22% Hispanic, 12% Chinese, and 53% female. Furthermore, 93% of Whites, 91% of Blacks, 31% of Hispanics, and 4% of Chinese were born in the U.S. A total of 6,428 study participants had complete data and were included in this analysis. Further description of the design of MESA appears elsewhere.20

Demographic and socioeconomic measures

Demographic information was obtained from standardized questionnaires. Race/ethnicity was self-reported based on questions modeled in the year 2000 U.S. Census and classified as non-Hispanic White, non-Hispanic African American, Hispanic, and Chinese. Two dimensions of SES (Socioeconomic Status) were measured: education and income. Education was self-reported as the highest level of school completed in eight categories ranging from no school to graduate/professional degree. Study participants reported their total combined family income in 13 categories ranging from less than $5,000 to $100,000 or more. We also assigned continuous educational attainment levels and income in dollars for each category of education and income by assigning the midpoint response value of each of the original categories to each study participant.

CVD risk factors

Smoking status was self-reported and categorized as current smoker/non-smoker. Weight and height were measured using a balance beam scale and vertical ruler. Participants wore light clothing and no shoes. The formula for BMI was weight (kg) divided by height (m) squared. Three resting blood pressure readings were obtained using an automated oscillometric sphygmomanometer (Dinamap, Criticon, Inc., Florida). The mean of the second and third measurements were analyzed. Hypertension was defined according to JNC VI (1997) criteria as diastolic blood pressure ≥90 mm Hg, or systolic blood pressure ≥140 mm Hg, or prior diagnosis of hypertension and current use of antihypertensive medications.21 Glucose levels were measured from blood samples after 12-hour fast using CDC (Centers for Disease Control and Prevention)-standardized methods. Diabetes was defined according to 2003 American Diabetes Association criteria as fasting glucose ≥126 mg/dl, use of insulin or oral hypoglycemic agents, or self-report of physician diagnosis.22

Statistical analysis

We investigated the SES patterning of the risk factors by comparing means and proportions across income and educational categories within each racial/ethnic group. Linear and logistic regression models were utilized to estimate age-adjusted means and proportions. Trend tests across categories were conducted by including ordinal income and educational categories in the models as continuous variables.

Because there was no consistent evidence of thresholds and in order to compare the SES patterning of income and education we estimated mean differences and odds ratios per one standard deviation (SD) unit increase in SES using continuous measures of the variables in regression models. All analyses were stratified by gender because we were specifically interested in gender heterogeneities and because prior work has found evidence of difference in the social patterning in men and women.23 Heterogeneity by race/ethnicity in SES patterning by gender was investigated through qualitative comparisons of the patterns in different racial/ethnic groups and by including SES race/ethnicity interaction terms in gender-stratified models pooling across the four races/ethnic groups. We report overall P values for heterogeneity as well as P values comparing associations in each racial/ethnic group with that observed in Whites.


Table 1 shows selected sample characteristics by race/ethnicity and gender. The median age of the sample was 62 years with little variation across groups. In general, White women and men were more highly educated and had higher incomes than members of the other racial/ethnic groups. Hispanic and Chinese participants were most likely to be in the lowest education or income categories (over 50% of Hispanic women and men, and Chinese women had only a high school education, compared with 30% or less for the other sex-ethnicity groups; over 40% of Hispanic and Chinese participants had incomes below $25,000, compared with 36% of Black women and less than 25% in other gender-race/ethnicity groups). With the exception of education in Blacks (which was similarly distributed in women and men), women were more likely to be in the lower educational and income categories than men. Relative to national income data reported by the U.S. Census, our sample included a slightly lower percentage of Whites and Blacks and a slightly greater percentage of Hispanics with income below $25,000.24 Additionally, a smaller percentage of Chinese in the sample had incomes above $75,000 compared with Asians nationally.24 The Spearman’s rho for the correlation of education and income was .47 for Whites, .44 for Blacks, .43 for Hispanics, and .45 for Chinese.

Table 1

Among women, Blacks had the highest prevalence of hypertension (60.4%) and smoking (16.1%), and the largest mean BMI (Body Mass Index) (31.3 kg/m2). Chinese women had the lowest rates of smoking (1.2%) and the lowest BMI (23.9 kg/m2). Hispanic women had the highest prevalence of diabetes (16.2%) followed closely by Blacks (15.6%). White women had the lowest prevalence of hypertension (37.8%) and diabetes (4.3%). Among men, Blacks had the highest prevalence of hypertension (55.2%), diabetes (19.2%), and smoking (19.4%), and the highest mean BMI (28.8 kg/m2). Chinese men had the lowest rates of hypertension (34.9%) and smoking (9.7%) and the lowest BMI (24.0 kg/m2). Hispanic men had BMI and diabetes prevalence similar to those of Black men (BMI 28.7 kg/m2 and diabetes 18.5%). White men had the lowest prevalence of diabetes (7.1%) but other risk factors were generally comparable to those in other racial/ethnic groups.

Figures 12 show the age-adjusted risk factor prevalence rates and means by gender, SES, and race/ethnicity. In White and Black women, education and income were associated in approximately graded fashion with all the risk factors, with the lower SES groups having worse cardiovascular profiles (although the trend in smoking by education was not statistically significant in Black women). Among Hispanic women, income and education were also inversely associated with most of the risk factors (except smoking and hypertension for which point estimates suggested a pattern but trend tests were not statistically significant). Patterns were much less consistent in Chinese women.

Figure 1
Age-adjusted prevalence of cvd risk factors by race/ethnicity, gender, and education.[large star]
Figure 2
Age-adjusted prevalence of cvd risk factors by race/ethnicity, gender, and income. [large star]

Among White men, higher SES was associated with lower risk factor levels, although for all risk factors except smoking the socioeconomic patterning was stronger and more likely to be statistically significant for education than for income. In Black men, smoking was inversely associated with income and education, but there was only weak, inconsistent, and non-statistically significant socioeconomic patterning of diabetes and hypertension. Additionally, income was positively associated with BMI among Black men. In Hispanic men both SES measures were inversely associated with diabetes, but no other statistically significant associations were observed, although point estimates suggested as positive association of income with BMI. In Chinese men smoking was inversely associated with education, and BMI was positively associated with income.

Tables 2 and and33 show odds ratios (OR) of diabetes, hypertension, and smoking and mean differences in BMI per standard deviation (SD) increase in education and income for women and men respectively. Among women (Table 2), higher education (ed) and income (inc) were usually associated with better risk factor profiles. This pattern was clearest in White women (OR ed diabetes: 0.68, 95% CI (Confidence Interval) 0.46–1.00; OR ed hypertension: 0.66, 95% CI 0.55–0.80; OR ed smoking: 0.36, 95% CI 0.28–0.48, mean difference ed BMI: −1.51, 95% CI −2.00–−1.02) followed by Black women (OR ed diabetes: 0.75, 95% CI 0.59–0.94; OR ed hypertension: 0.75, 95% CI 0.61–0.92; OR ed smoking: 0.77, 95% CI 0.61–0.99, mean difference BMI: −1.21, 95% CI −1.76–−0.66). Among Hispanic women similar inverse associations were observed for diabetes (OR ed diabetes 0.72, 95% CI 0.60–0.86) but no statistically significant associations were observed for the other outcomes. Very similar results regarding racial/ethnic differences in the associations were observed for income, with the exception that greater income was associated with lower BMI in Hispanic women (mean difference BMI −0.78 95% CI −1.47, −0.09). None of the associations of income or education with the outcomes were statistically significant in Chinese women, but results were highly variable due to smaller sample size and low prevalence of some outcomes.

Table 2
Table 3

In women overall, statistical interactions between education and race/ethnicity were statistically significant (overall P for interaction <.05, Table 2) for three of the four risk factors (diabetes being the exception) indicating evidence of heterogeneity in the educational patterning by race/ethnicity. More specifically, for smoking, hypertension, and BMI, the inverse patterning was significantly weaker in Chinese and Hispanic women than in White women (all six P for interactions comparing each race/ethnic group with Whites <.05, not shown in table). The only statistically significant difference between Black and White women was observed for smoking, with weaker patterns among Black than White women (P for interaction comparing Blacks with Whites <.05). In contrast, there was no statistical evidence of heterogeneity in the effects of income by race/ethnicity (all overall P for interaction >.05) although Chinese women showed associations of income with smoking and BMI that differed significantly from those observed in Whites (P for interaction Chinese vs. Whites <.05).

Among men (Table 3), income and education were generally inversely associated with risk factors in Whites but patterns differed in magnitude and even in direction for the other races/ethnic groups. All of the outcomes showed the strongest inverse educational patterning among White men (OR ed diabetes: 0.76, 95% CI 0.57–1.01; OR ed hypertension: 0.77, 95% CI 0.65–0.92; OR ed smoking: 0.51, 95% CI 0.40–0.66, mean difference BMI: −0.60, 95% CI, −0.93–−0.27). Education was also inversely associated with diabetes across all racial/ethnic groups, although the association in Chinese men was not statistically significant (OR ranging from 0.76 to 0.85). In contrast education was not inversely associated with hypertension or BMI in non-White men, and education was inversely associated with smoking in Blacks (OR: 0.64, 95% CI 0.50–0.82) and Chinese (OR: 0.59, 95% CI 0.44–0.80) but not Hispanics.

Less consistent patterns were observed for income in men. Associations of income with diabetes were strongest in Hispanic men (OR: 0.64, 95% CI 0.48–0.87) and weakest in White men (OR: 0.94, 95% CI 0.75–1.17) and Chinese men (OR: 1.27, 95% CI 0.94–1.71). Associations of income with hypertension were not statistically significant in any racial/ethnic group. Associations of income with smoking were strongest in Black men (OR: 0.57, 95% CI 0.46–0.72) and weakest in Chinese men (OR: 0.84, 95% CI 0.58–1.21). The strongest associations of income with BMI were observed in Black men (mean diff BMI: 0.48, 95% CI 0.11–0.85) and Chinese (mean diff BMI: 0.47, 95% CI 0.16–0.78) men, but the association was positive, that is higher income was associated with higher BMI. Only weak and non-statistically significant associations of income with BMI were observed in White and Hispanic men.

As in women, there was evidence of heterogeneity in the social patterning by education: for three of the four risk factors (the exception being diabetes) the overall interaction between education and race/ethnicity was statistically significant (p<.05). More specifically, the educational patterning of BMI in Black, Hispanic, and Chinese men differed significantly from that observed in White men (all three P for interaction comparing association in each race/ethnic group with that observed in Whites <.05) and the educational patterning of smoking and hypertension differed significantly in Hispanic and White men (p for interaction <.05). In men, the income patterning of diabetes and BMI also differed by race/ethnicity (both overall p<.05). Specifically, the BMI patterning by income in Black, Chinese, and Hispanic men differed significantly from that observed in Whites and the income patterning of diabetes differed significantly in Hispanic and White men (all P values for interactions for each race/ethnic group vs. Whites <.05).

There was some (albeit limited) evidence that education was a stronger predictor of outcomes in Whites than in other racial/ethnic groups. For example, the inverse associations of education with the risk factors were systematically stronger in White women than in other women whereas findings for income were more variable. Approximately similar patterns were observed in men for diabetes, hypertension, and smoking—all three were consistently stronger predictors in White men than in other racial/ethnic groups, whereas patterns for income were more variable. Additionally, BMI was clearly negatively associated with education in White men, whereas it was weakly positively associated (or unassociated) with education in the other races/ethnic groups.


To our knowledge this is one of few studies to investigate heterogeneity in SES gradients for CVD risk factors in a diverse U.S. sample. We found evidence of important heterogeneity in the associations of SES with the risk factors by race/ethnicity and gender. Specifically, inverse socioeconomic gradients in hypertension, diabetes, smoking, and BMI were observed in White and Black women but associations were weaker or absent in Hispanic and Chinese women (except in the case of diabetes for Hispanic women). Even greater heterogeneity in social patterning of risk factors was observed in men. In White men, all four risk factors were inversely associated with socioeconomic position, although often associations were only present or were stronger for education than for income. In contrast, with the exception of diabetes among Hispanic men and smoking in Chinese men, the inverse socioeconomic patterning was much less consistent in men of other races/ethnic groups. In contrast to the inverse associations of education with BMI in White men, there was some suggestion that higher SES was associated with higher rather than lower BMI in non-White men. There was clear statistical evidence of heterogeneity in the SES patterning for education in both men and women, and for all the risk factors investigated except diabetes.

These results are important for two reasons. First they illustrate that the social patterning is not invariant by race/ethnicity and gender. Further research on why this is the case could shed important light on the determinants of cardiovascular risk in different racial/ethnic groups. In addition, greater understanding of this heterogeneity could also enhance our understanding of the many processes through which social factors affect cardiovascular disease. The findings also have implications for analyses that attempt to examine the persistence of racial/ethnic differences after adjusting for SES. The differential association of SES with risk factors implies that racial/ethnic differences are also modified by SES and that efforts to estimate SES-adjusted race/ethnic differences must account for these statistical interactions. Finally, these findings also have implications for the targeting and tailoring of preventive interventions by socioeconomic, racial/ethnic, and gender groups.

The inverse SES patterns in CVD risk factors we observed for Whites were consistent with findings from numerous prior studies.9 Also consistent with prior work, we often found stronger inverse social patterning of CVD risks in women than in men.23 Our results were similar to those of other studies showing inverse social patterning of cardiovascular risk factors in Black women but more variable and sometimes even opposite associations in Black men. For example, at least one previous study of U.S. Blacks has shown that low SES was associated with high BMI in Black women but was not as clearly associated with BMI in Black men.25 Other work has also documented either inconsistent or positive associations of SES with cardiovascular risk among Black men in some contexts.26,27

Relatively few studies have examined social patterning of CVD risk factors in U.S. Hispanics. Previous studies of U.S. Hispanics have shown poorer cardiovascular profiles for obesity, diabetes, hypertension, and smoking among those with lower education although educational differentials are often less consistent and weaker than in other racial/ethnic groups.28,29 Our findings of less consistent SES patterning in Hispanic women and men compared with White women and men is therefore consistent with prior work. Similar to our results, at least one other study has reported more consistent socioeconomic patterning of risk factors among Hispanic women than among Hispanic men.25

Most of the previous research on the SES patterning of CVD risk factors among Chinese groups has shown positive associations of SES with risk factors among people living in China, no gradients or some positive gradients among foreign-born Chinese in the U.S., and inverse gradients among U.S.-born Chinese.3032 Most of the Chinese participants in our study (over 90%) were foreign-born, which may explain the weak or absent SES patterning that we observed. However, results for Chinese are strongly limited by small sample size which resulted in very wide confidence intervals.

A number of factors could explain the observed heterogeneity in SES patterning by race/ethnicity and gender. These include differential material and psychosocial consequences of greater income or education in different groups, factors related to recent immigration, and selection factors. The impact of greater income and education for health may differ by race/ethnicity due to a variety of factors such as discrimination, segregation, and inter-generational life-course processes.33 For example, in the U.S., Whites with higher incomes and education may be more likely than high-SES persons from other groups to benefit from better material living conditions and work environments, and live in less stressful and more health-enhancing environments. They may also be more likely to have lived in advantaged households as children. This may explain why the socioeconomic patterning was generally stronger in Whites than in other races/ethnic groups.

Black women in this cohort may have been especially affected by public policies, such as civil rights legislation that likely allowed them to take advantage of expanded higher education programs, the opening of service-sector jobs, and associated wage gains with consequent gains for physical and mental health.34 This may have allowed more educated and higher-income Black women to derive more benefits from their higher social standing compared to less educated and lower income Black women, resulting in the observed strong SES patterning of CVD risks in Black women. A striking caveat to this is that despite the better CVD risk factor profile in higher-SES compared with lower-SES Black women, our data show that for all risk factors except smoking that the most advantaged Black women had similar or greater adverse risk profiles than the most disadvantaged White women.

The positive associations of income with BMI in non-White men could also reflect differential material and psychosocial correlates of SES by gender and race/ethnicity. For example, the positive income gradients in Black men for BMI may be indicative of excessive stress and/or high effort coping among higher-income earners, resulting in poorer lifestyle choices.35 Alternatively, the adverse BMI profiles among high-income Black men in our sample may also be indicative of a greater adoption of a “rich” diet (resulting in greater caloric and fat consumption) reflecting their higher status, particularly in light of the history of oppression, racial discrimination, and minimal post–civil rights advancement experienced by this group in the U.S.36,37

Immigration history may also play a role in the heterogeneity that we observed. The weaker and less consistent socioeconomic patterning in U.S. Chinese and Hispanics, generally, may partly reflect socioeconomic patterning of CVD in places of origin. For example, positive SES gradients in CVD risk factors have been documented in Puerto Rico, China, and Mexico.3840 The health consequences of greater income and education may differ for these groups due to differential material and psychosocial correlates for immigrants in their countries of origin that they bring with them to their host countries when they immigrate.

Selection factors could also play a role. For example, lower-SES migrants may be more strongly selected for healthier CVD profiles than higher-SES migrants.28 Differential study participation rates by SES in non-White groups could have resulted in biased estimates if low SES participants are selected to be especially healthy. However, because we have no information on the health status of nonparticipants the role of selection remains highly speculative.

Few studies have investigated differences in socioeconomic patterning of CVD risk factors by both education and income. Winkleby et al. reported education to be a stronger and more consistent predictor of CVD risk than income or occupation among U.S. Whites.41 We found racial/ethnic differences in the SES patterning to be more pronounced for education than for income with the strongest inverse associations between education and the CVD risk factors in Whites. This is consistent with prior work showing greater income returns on education for Whites than for other races/ethnic groups.12,13

Strengths of this study include the large multiethnic population sample and the standardized measures of risk factors. The main limitations include small within-group sample sizes for some gender/race groups, as well as the cross-sectional nature of the analyses. Some null associations in non-Whites may have possibly caused by lack of statistical power, and fluctuations in the direction of some of the associations may also have been caused by chance. The sample was not intended to be representative of race/ethnicity in the U.S., although the socioeconomic distribution was not very different from what is observed in national data. Although our results were often consistent with results reported in other samples, it is possible that nationally representative samples could yield different results. In particular, the fact that Hispanics and Chinese were not recruited from all sites possibly limits the generalizability for these racial/ethnic groups further.

Additionally, there are undoubtedly limitations to our assessment of SES. We had imperfect measure of education and income which may have limited our ability to detect statistically significant heterogeneity in associations. Our income measure was limited in that it was categorical and did not allow fine adjustment for family size. We did examine correlations between per capita income (calculated based in the category midpoints) and the categorical measure of income we used and found high correlations, which suggest that family size would not have a large impact on our results although finer measures of income would have allowed a much more accurate adjustment for family size.

Our findings clearly point to the fact that the CVD risk factors are patterned by SES within racial/ethnic and gender groups but that important variation exists in the strength and direction of such patterning. Understanding why this heterogeneity exists may provide important clues on the causes of CVD and the reasons for racial/ethnic disparities. This heterogeneity also must be considered when analysts attempt to estimate racial/ethnic differences after statistical adjustment for SES.


This research was supported by the National Institutes of Health, National Center for Minority Health and Health Disparities, through the Center for Integrative Approaches to Health Disparities at the University of Michigan by grant P60 MD002249. MESA is supported by contracts N01-HC-95159 through N01-HC-95165 and N01-HC-95169 from the National Heart, Lung, and Blood Institute. The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions.

Contributor Information

Shawn Boykin, Center for Integrative Approaches to Health Disparities in the Department of Epidemiology at the University of Michigan (UM)

Ana V. Diez-Roux, Center for Integrative Approaches to Health Disparities and the Center for Social Epidemiology in Population Health in the Department of Epidemiology at UM.

Mercedes Carnethon, Department of Preventative Medicine at the Feinberg School of Medicine at Northwestern University.

Sandi Shrager, Department of Biostatistics at the University of Washington.

Hanyu Ni, Division of Epidemiology and Clinical Applications at the National Heart Lung Blood Institute, National Institutes of Health.

Melicia Whitt-Glover, Gramercy Research Group in Winston-Salem, NC.


1. Wong MD, Shapiro MF, Boscardin WJ, et al. Contribution of major diseases to disparities in mortality. N Engl J Med. 2002 Nov 14;347(20):1585–92. [PubMed]
2. Kramer H, Han C, Post W, et al. Racial/ethnic differences in hypertension and hypertension treatment and control in the multi-ethnic study of atherosclerosis (MESA) Am J Hypertens. 2004 Oct;17(10):963–70. [PubMed]
3. McBean AM, Li S, Gilbertson DT, et al. Differences in diabetes prevalence, incidence, and mortality among the elderly of four racial/ethnic groups: Whites, Blacks, Hispanics, and Asians. Diabetes Care. 2004 Oct;27(10):2317–24. [PubMed]
4. Winkleby MA, Kraemer HC, Ahn DK, et al. Ethnic and socioeconomic differences in cardiovascular disease risk factors: findings for women from the Third National Health and Nutrition Examination Survey, 1988–1994. JAMA. 1998 Jul 22–29;280(4):356–62. [PubMed]
5. Williams DR, Mohammed SA, Leavell J, et al. Race, socioeconomic status, and health: complexities, ongoing challenges, and research opportunities. Ann N Y Acad Sci. 2010 Feb;1186:69–101. [PMC free article] [PubMed]
6. Mujahid MS, Diez Roux AV, Cooper RC, et al. Neighborhood stressors and race/ethnic differences in hypertension prevalence (the multi-ethnic study of atherosclerosis) Am J Hypertens. 2010 Sep 16; [PMC free article] [PubMed]
7. Thomas AJ, Eberly LE, Davey Smith G, et al. Race/ethnicity, income, major risk factors, and cardiovascular disease mortality. Am J Public Health. 2005 Aug;95(8):1417–23. Epub 2005 Jul 7. [PMC free article] [PubMed]
8. Whitty CJ, Brunner EJ, Shipley MJ, et al. Differences in biological risk factors for cardiovascular disease between three ethnic groups in the Whitehall II study. Atherosclerosis. 1999 Feb;142(2):279–86. [PubMed]
9. Kaplan GA, Keil JE. Socioeconomic factors and cardiovascular disease: a review of the literature. Circulation. 1993 Oct;88(4 Pt 1):1973–98. [PubMed]
10. Morgenstern H. The changing association between social status and coronary heart disease in a rural population. Soc Sci Med. 1980 May;14A(3):191–201. [PubMed]
11. Marmot MG. Socio-economic and cultural factors in ischaemic heart disease. Adv Cardiol. 1982;29:68–76. [PubMed]
12. Ashraf J. Differences in returns to education: an analysis by race. Am J Econ Sociol. 1994 Jul;53(3):281–90.
13. Carliner G. Returns to education for Blacks, Anglos, and five Spanish groups. J Hum Resour. 1976 Spring;11(2):172–84.
14. Borrell LN, Jacobs DR, Jr, Williams DR, et al. Self-reported racial discrimination and substance use in the Coronary Artery Risk Development in Adults Study. Am J Epidemiol. 2007 Nov 1;166(9):1068–79. Epub 2007 Aug 13. [PubMed]
15. Williams DR. Race, socioeconomic status, and health. The added effects of racism and discrimination. Ann N Y Acad Sci. 1999;896:173–88. [PubMed]
16. Williams DR, Yan Y, Jackson JS, et al. Racial differences in physical and mental health: socioeconomic status, stress and discrimination. J Health Psychol. 1997 Jul 1;3(2):335–51. [PubMed]
17. Kreiger N. Embodying inequality: a review of concepts, measures, and methods for studying health consequences of discrimination. Int J Health Serv. 1999;29(2):295–352. [PubMed]
18. DeNavas-Walt C, Proctor BD, Smith J. Income, poverty, and health insurance coverage in the United States: 2006. Washington, DC: U.S. Census Bureau; 2007. Available at: http://www.census.gov/prod/2007pubs/p60-233.pdf.
19. Jackson SA, Anderson RT, Johnson NJ, et al. The relation of residential segregation to all-cause mortality: a study in Black and White. Am J Public Health. 2000 Apr;90(4):615–7. [PMC free article] [PubMed]
20. Bild DE, Bluemke DA, Burke GL, et al. Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol. 2002 Nov 1;156(9):871–81. [PubMed]
21. National High Blood Pressure Education Program. The Sixth Report of the Joint National Committee on the Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC VI) Bethesda, MD: National Heart, Lung, and Blood Institute; 1997. Available at: http://www.ncbi.nlm.nih.gov/bookshelf/br.fcgi?book=hbp. [PubMed]
22. Genuth S, Alberti KG, Bennett P, et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care. 2003 Nov;26(11):3160–7. [PubMed]
23. Robbins JM, Vaccarino V, Zhang H, et al. Socioeconomic status and type 2 diabetes in African American and non-Hispanic White women and men: evidence from the Third National Health and Nutrition Examination Survey. Am J Public Health. 2001 Jan;91(1):76–83. [PMC free article] [PubMed]
24. U.S. Census Bureau. Current Population Survey: Annual Social and Economic (ASEC) Supplement Survey, 2007. Ann Arbor, MI: Inter-university Consortium for Political and Social Research; 2008.
25. Sanchez-Vaznaugh EV, Kawachi I, Subramanian SV, et al. Do socioeconomic gradients in body mass index vary by race/ethnicity, gender, and birthplace? Am J Epidemiol. 2009 May 1;169(9):1102–12. Epub 2009 Mar 18. [PubMed]
26. Diez-Roux AV, Nieto FJ, Muntaner C, et al. Neighborhood environments and coronary heart disease: a multilevel analysis. Am J Epidemiol. 1997 Jul 1;146(1):48–63. [PubMed]
27. Dressler WW. Lifestyle, stress, and blood pressure in a southern Black community. Psychosom Med. 1990 Mar–Apr;52(2):182–98. [PubMed]
28. Kimbro RT, Bzostek S, Goldman N, et al. Race, ethnicity, and the education gradient in health. Health Aff (Millwood) 2008 Mar–Apr;27(2):361–72. [PubMed]
29. Goldman N, Kimbro RT, Turra CM, et al. Socioeconomic gradients in health for White and Mexican-origin populations. Am J Public Health. 2006 Dec;96(12):2186–93. Epub 2006 Oct 31. [PMC free article] [PubMed]
30. Socio-economic status and risk factors for cardiovascular disease: a multicentre collaborative study in the International Clinical Epidemiology Network (INCLEN) The INCLEN Multicentre Collaborative Group. J Clin Epidemiol. 1994 Dec;47(12):1401–9. [PubMed]
31. Tian HG, Nan Y, Liang XQ, et al. Relationship between serum lipids and dietary and non-dietary factors in a Chinese population. Eur J Clin Nutr. 1995 Dec;49(12):871–82. [PubMed]
32. Lauderdale DS, Rathouz PJ. Body mass index in a U.S. national sample of Asian Americans: effects of nativity, years since immigration and socioeconomic status. Int J Obes Relat Metab Disord. 2000 Sep;24(9):1188–94. [PubMed]
33. Williams DR, Neighbors HW, Jackson JS. Racial/ethnic discrimination and health: findings from community studies. Am J Public Health. 2003 Feb;93(2):200–8. [PMC free article] [PubMed]
34. Kaplan G, Ranjit N, Burgard S. Lifting gates, lengthening lives: did civil rights policies improve the health of African-American women in the 1960’s and 1970’s? In: Schoeni RF, House JS, Kaplan GA, et al., editors. Making Americans healthier: social and economic policy as health policy. New York, NY: Russell Sage Foundation; 2008.
35. James SA, Keenan NL, Strogatz DS, et al. Socioeconomic status, John Henryism, and blood pressure in Black adults. The Pitt County Study. Am J Epidemiol. 1992 Jan 1;135(1):59–67. [PubMed]
36. Kantor H, Brenzel B. Urban education and the “truly disadvantaged”: the historical roots of contemporary crisis, 1945–1990. In: Katz M, editor. The “underclass” debate: views from history. Princeton, NJ: Princeton University Press; 1993.
37. Variyam JN. Patterns of caloric intake and body mass index among U.S. adults. Food Review. 2002;25(3):16–20.
38. Yu Z, Nissinen A, Vartiainen E, et al. Socio-economic status and serum lipids: a cross-sectional study in a Chinese urban population. J Clin Epidemiol. 2002 Feb;55(2):143–9. [PubMed]
39. Sorlie PD, Garcia-Palmieri MR. Educational status and coronary heart disease in Puerto Rico: the Puerto Rico Heart Health Program. Int J Epidemiol. 1990 Mar;19(1):59–65. [PubMed]
40. Fernald LC. Socio-economic status and body mass index in low-income Mexican adults. Soc Sci Med. 2007 May;64(10):2030–42. Epub 2007 Mar 21. [PMC free article] [PubMed]
41. Winkleby MA, Jatulis DE, Frank E, et al. Socioeconomic status and health: how education, income, and occupation contribute to risk factors for cardiovascular disease. Am J Public Health. 1992 Jun;82(6):816–20. [PMC free article] [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...