Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Ann Epidemiol. Author manuscript; available in PMC 2011 August 1.
Published in final edited form as:
PMCID: PMC2901883

Socio-economic and Ethnic Disparities in Cardiovascular Risk In the United States, 2001-2006



To quantify socioeconomic status and ethnic differences in risk for coronary heart disease (CHD) accrued from major risk factors, in the United States (US).


Data came from the National Health and Nutrition Examination Survey 2001-2006. Outcomes examined were a) 10-year risk for CHD events as predicted by the National Cholesterol Education Program Adult Treatment Panel III 2004 Updated guidelines, and b) the prevalence of the metabolic syndrome and overt diabetes mellitus (a CHD risk-equivalent).


Strong inverse socioeconomic gradients with risk were present in all race/ethnicity groups except foreign-born Mexican American men, and were attenuated by controls for physical activity, smoking, and abdominal obesity. In contrast, race/ethnicity disparities were seen in some but not all socioeconomic strata, with some Non-Hispanic Blacks and US-born Mexican Americans having higher risk and some Foreign-born Mexican Americans having lower risk.


Disparities in cardiovascular risk in the United States are primarily related to SES, and less to race/ethnicity. Socioeconomically disadvantaged individuals should be targeted for lifestyle counseling and early screening for risk factors, regardless of race/ethnicity, to reduce social disparities in cardiovascular outcomes.

Keywords: Socioeconomic status, ethnic differences, Framingham risk score, metabolic syndrome


The incidence of cardiovascular disease in the United States (US) and other high-income countries is higher in low socio-economic status (SES) and minority ethnic groups (1,2,3,4,5), with SES disparities reflecting not just a threshold effect of poverty versus non-poverty, but rather a risk gradient across the SES range (2,6). Reduction of these disparities has been a focus of interest for policy makers (7,8), and requires the early detection and management of risk factors in underprivileged groups. Indeed, low SES and minority ethnic groups do report higher values of risk factors for cardiovascular disease (1,9,10).

Some studies, however, suggest that SES gradients in cardiovascular risk factors may be diminished in minority groups (11,12,13,14), possibly because of racial discrimination and residential segregation (15,16,17), and greater psychological distress (18) at all SES levels. If SES associations with risk indeed do vary by race/ethnicity, then in any study of ethnic disparities, SES effects cannot be completely controlled simply by including SES indicators as covariates in regression (19,20). It is therefore, not clear if ethnic differences in cardiovascular risk primarily reflect SES differences, or if there are SES-independent race/ethnicity associations with risk; the two scenarios have different implications for health policy.

Moreover, few studies have examined SES / ethnic differences in global or overall cardiovascular risk accrued from all major risk factors. When SES /ethnic disparities in individual risk factors are of varying size and/or are in opposite directions (5,21), their combined effect can only be assessed by the use of global risk measures. Current clinical guidelines assess global risk for coronary heart disease (CHD) events using the Framingham risk score, the metabolic syndrome, and history of CHD and CHD risk-equivalent conditions (such as diabetes mellitus that confer equivalent risk) (22,23,24). Accordingly, in this work, we studied ethnic and SES disparities in the US population in a) the 10-year risk for CHD events, as assessed by the National Cholesterol Education Program (NCEP) Adult Treatment Panel III adaptation of the Framingham risk score (22,25), and b) the prevalence of diabetes mellitus and metabolic syndrome, as defined by the NCEP (22) and the European Group for the Study of Insulin Resistance (EGIR) (26). In particular, our objectives were to answer the following questions:

  • Q1. Are there SES gradients in global CHD risk and metabolic syndrome /diabetes prevalence, and do the SES gradients differ by race/ethnicity (i.e., do minority groups not get the same benefit from climbing the SES ladder as the majority ethnic group)?
  • Q2. Are there race/ethnicity differences in global CHD risk and metabolic syndrome/diabetes prevalence after adequately controlling for differences in SES?
  • Q3. Are SES and race/ethnicity associations with global CHD risk and metabolic syndrome/diabetes prevalence explained by differences in health behaviors and abdominal obesity (a surrogate for both diet and exercise)?


Data for this study came from the National Health and Nutrition Examination Survey (NHANES), conducted between 2001 and 2006, of a nationally representative sample of the US population (27).

Study Sample

Of the 31,509 participants in NHANES 2001-2006, we excluded those under the age of 20 (n=16,078), pregnant women (n=886), individuals who missed the clinical examination (n=843), and those missing SES data (n=24). For the CHD risk analyses, we additionally excluded 1,524 individuals whose global CHD risk could not be classified; the resulting sample size was 12,154 (83.6% of those age-eligible and not pregnant). For the metabolic risk analyses, we excluded 7,970 individuals whose metabolic syndrome / diabetes data was missing (primarily because they had not fasted 6 hours or more before the blood draw); the final sample size for metabolic risk analysis was 5,648. Comparison of the analytic samples with the rest of the non-pregnant NHANES sample 20 years or older showed that they were representative of the complete NHANES sample (Table 1).

Table 1
Descriptive Characteristics of Study Samples from National Health and Nutrition Examination Survey (NHANES) 2001-2006, at least 20 years of age, and not pregnant; weighted to be nationally representative


Demographic, biological, and health data were collected at a home interview, followed by an examination and blood draw, either at home or a mobile examination center. Age, gender, race/ethnicity, education, and income data were obtained from standard adult questionnaires.

Educational attainment was classified as less than high school, complete high school, or more than high school education. Poverty Income Ratio (PIR), the ratio of household income to the poverty level specific to area of residence and household size (28), was recorded. Since educational attainment does not provide the same returns with respect to income in all race/ethnicity groups (15), we combined income and education to create three SES categories: We defined low SES as education less than high school and PIR <2, and high SES as education more than high school and PIR≥2. Everyone else was classified as middle SES, which included those with a) just high school education regardless of PIR, b) less than high school education but PIR≥2, and c) more than high school education but PIR<2.

Race/ethnicity was classified as Non-Hispanic White, Non-Hispanic Black, US-born Mexican American, Foreign-born Mexican American, and other. We distinguished between US-born and Foreign-born Mexican Americans because health risks in the 2 groups appear to be quite different, even at comparable SES (29,30,31).

Self reported medical history, physical activity (moderate / vigorous activity vs. none), current smoking, and medication data were obtained from the NHANES adult questionnaire. Prescription medication information was verified from medication bottle review. Blood pressure and waist circumference were measured at the examination. Blood samples were assayed for serum levels of C-reactive protein (CRP), total cholesterol, HDL cholesterol, fasting low density lipoprotein (LDL) cholesterol, fasting glucose, and fasting triglycerides. The last three assays were obtained only for those participants who had fasted at least 6 hours.

Risk Assessment

The two primary outcomes were global CHD risk and metabolic syndrome /diabetes prevalence. We created six CHD risk groups (very low risk to very high risk) based on the NCEP Adult Treatment Panel III guidelines 2004 update (22,25). Individuals with prevalent CHD (based on self report of CHD, angina, or myocardial infarction) or CHD-risk equivalent condition (diabetes mellitus or self reported stroke) were classified as high risk. Prevalence of diabetes was assessed from self-report (if not diagnosed during pregnancy) and/or fasting blood glucose > 126 mg/dL (7 mmol/L). Those with prevalent CHD and one or more of diabetes mellitus, metabolic syndrome (assessment described below), and current smoking, were classified as very high risk (the highest category of risk in the NCEP 2004 Updated Guidelines). Ten-year global risk for CHD in individuals without prevalent CHD or CHD-risk-equivalent condition was estimated from the Framingham risk score, a gender-specific measure based on age, current smoking, total cholesterol, HDL cholesterol, blood pressure, and anti-hypertensive medication use. We scored those 80 years and older like 70-79 year olds (the oldest group in the scoring system). Based on the predicted global CHD risk, participants without prevalent CHD or CHD-risk-equivalent were classified as either very low risk (if predicted 10-year risk <1%); low risk (predicted 10-year risk 1-9%); moderate risk (if predicted 10-year risk 10-19% and without metabolic syndrome, not currently smoking, and serum CRP less than 3 mg/L), moderate plus risk (if predicted 10-year risk 10-19% and one or more of current smoking, metabolic syndrome, or CRP>=3 mg/L); or high risk (predicted 10-year risk ≥ 20%), consistent with risk categories used in clinical guidelines for cardiovascular risk management (22,23,24,25). For the purposes of this risk categorization, individuals with missing data on metabolic syndrome, diabetes, or CRP were assumed to not meet that risk criterion. Misclassification resulting from this imputation is limited to at most one category (e.g., classified moderate instead of moderate-plus, or high instead of very high).

We also created three metabolic syndrome /diabetes groups: Low risk: no metabolic syndrome and no diabetes mellitus; Moderate risk: metabolic syndrome but without diabetes mellitus, and High risk: diabetes mellitus. Based on the NCEP and EGIR definitions (22,26), metabolic syndrome was defined as having three or more of the following conditions: 1) Waist circumference >102 cm (40 inches) for men, >89 cm (35 inches) for women; 2) Fasting serum triglycerides>= 150 mg/dL (1.7 mmol/L); 3) HDL cholesterol< 40 mg/dL (1 mmol/L) for men, < 50 mg/dL (1.3 mmol/L) for women; 4) Blood pressure >=130mmHg systolic or >=85 mmHg diastolic, 5) Fasting blood glucose >= 110 mg/dL (6.1 mmol/L). Because of the high concordance between high fasting triglycerides and high waist circumference, we imputed one condition from the other (i.e., imputed high on missing condition if high on the other, low on missing condition if low on other) in 58 participants who were missing either triglycerides or waist circumference, to allow them to be classified.

Statistical Analyses

All analyses of ethnic and SES differences in risk were age-standardized; i.e., the age distribution of each predictor group (each SES group and each race/ethnicity group in unadjusted analyses, SES groups within race strata in race-stratified analyses, and race/ethnicity groups within SES strata in SES-stratified analyses) was standardized to the age distribution of the US in the 2000 census. Cumulative logistic regression models were then fit to determine the cumulative odds ratio (cumulative OR) for CHD risk and metabolic risk categories, with respect to SES and race/ethnicity, separately in each gender. Cumulative logistic regression for ordinal outcomes (such as the ordered CHD risk and metabolic risk categories) models the log of the cumulative odds at every outcome level (which are the odds of being in that or a higher outcome level versus all lower levels) as a common, single linear function of predictors plus level-specific intercepts. The cumulative OR for a predictor is the average effect of the predictor on cumulative odds at all outcome levels. In addition, we also estimated the ordinary odds ratio (OR) for the highest risk categories (very high risk for the 6-level global CHD risk outcome and diabetes mellitus for the 3-level metabolic risk outcome) versus all other risk categories, using ordinary (binary) logistic regression. Analyses were stratified by gender because CHD and metabolic risk categorizations were somewhat different in men and women (different thresholds for HDL, waist circumference, and different scoring systems for 10-year CHD risk), and because SES is felt to confer different benefits in men and women (3,32).

To quantify SES associations, gender and race/ethnicity-stratified, age-standardized models were fit with SES as primary predictor, and to assess race/ethnicity differences in risk, gender and SES-stratified, age-standardized models were fit with race/ethnicity as primary predictor. We chose to conduct race/ethnicity analyses stratified by SES and to conduct SES analyses stratified by race/ethnicity to minimize residual confounding, because of the preponderance of evidence that SES-risk associations vary by race/ethnicity (11,12,13,14). Given the existence of SES-race/ethnicity interactions, one cannot completely control for SES in a race/ethnicity analysis simply by including SES as a covariate in regression. Similarly, simply adding race/ethnicity to an SES regression model will not eliminate confounding by race/ethnicity.

To assess the role of lifestyle and central obesity in risk disparities, we added current smoking, physical activity (any moderate or vigorous activity vs. neither) and high waist circumference (>40 inches for men, >35 for women) to the models. We chose to study the role of abdominal obesity rather than generalized obesity, because of the former’s stronger associations with metabolic and cardiovascular risk (22,26).

Stata version 9 was used for all analyses. CHD risk analyses were weighted by NHANES exam weights, and metabolic risk analyses were weighted by NHANES fasting sample weights, to account for sampling probability and non-response. To account for clustering within primary sampling units (because of the survey design), we used Stata’s cluster option and robust variance estimation.


The analytic samples were older, more likely to be Non-Hispanic White and from higher socioeconomic strata, and had better biological profiles than the rest of the NHANES sample (See Table 1). Median age in the analytic samples was 45 years, 49% were male, 73% were Non-Hispanic White, 18% had less than high school education, median PIR was 3.1 and 12% had PIR less than 1, 11-12% fell in the low SES category, and 45% were classified high SES. There were marked SES differences between race/ethnicity groups: SES was highest in Non-Hispanic Whites and lowest in Foreign-born Mexican Americans. There were also marked differences by race/ethnicity in lifestyle choices and prevalence of central obesity (See Table 2). Among women, prevalence of both former smoking and current smoking was highest in Non Hispanic White women and lowest in Foreign-born Mexican American women (Table 2). Among men, current smoking prevalence was not very different between race/ethnicity groups, but Non-Hispanic Black men were the least likely to be former smokers. Non-Hispanic White men and women were also the most likely to be physically active (at moderate or vigorous level) and Foreign-born Mexican American men and women least likely (Table 2). On the other hand, central obesity prevalence was lowest in Foreign-born Mexican American men and women, and highest in Non-Hispanic White and US-born Mexican American men and Non-Hispanic Black and US-born Mexican American women (Table 2).

Table 2
Socioeconomic Status, Lifestyle Choices, and Central Obesity by Race/Ethnicity and Gender, United States 2001-2006

In unadjusted (but age-standardized) analyses, both education and income individually, and the combined SES variable had strong inverse relationships with global CHD risk and metabolic risk (data not shown). However, tests of interaction in gender-stratified, logistic regression models indicated that race/ethnicity modified these SES associations in both men and women: For instance, (age-standardized) SES associations with the odds of very high CHD risk (the highest CHD risk group) were significantly weaker in US-born Mexican American men (p=0.005) and stronger in foreign-born Mexican American women (p=0.02) than in the majority group (Non-Hispanic White). Similarly, SES associations with metabolic risk were weaker in both US-born Mexican men (p=0.1) and Non-Hispanic Black men (p=0.07) than in Non-Hispanic White men, and SES associations with overt diabetes (the highest metabolic risk group) were weaker in foreign-born Mexican American men (p=0.08) but stronger in foreign-born Mexican American women (p=0.09) than in Non-Hispanic Whites.

In analyses stratified by gender and race/ethnicity, inverse SES gradients were seen with global CHD risk in men and women from all race/ethnicity groups, with one notable exception: There was no SES gradient in foreign-born Mexican American men. Comparing low SES to high SES, cumulative OR for estimated 10-year CHD risk levels ranged from 1.4 to 2.8 and OR for being in the highest risk group ranged from 2.9 to 5.8 (Table 3a). After adjusting for lifestyle factors (physical activity and smoking) and central obesity, the magnitude of the SES-risk gradient diminished in all strata – See Table 3a.

Table 3a
Socioeconomic Status (SES) Differences in Estimated 10-year CHD Risk (Odds Ratios Compared to High SES, with 95% Confidence Intervals), Stratified by Race/Ethnicity, Before and After Adjusting for Physical Activity, Current Smoking, and Central Obesity; ...

In women from three of the four race/ethnicity groups and in non-Hispanic White men, SES also had strong inverse associations with metabolic syndrome / diabetes prevalence. The strongest SES effect was in foreign-born Mexican American women: Cumulative OR 2.5 (95% confidence interval (CI): 1.5 to 4.1) and diabetes OR 5.0 (95% CI: 2.2 to 11.0), for low SES compared to middle and high SES. Adjusting for physical activity and smoking only modestly attenuated these SES gradients – See Table 3b.

Table 3b
Socioeconomic Status (SES) Differences in Metabolic Risk (Odds Ratios Compared to High SES, with 95% Confidence Intervals), Stratified by Race/Ethnicity, Before and After Adjusting for Physical Activity and Current Smoking; United States 2001-2006

In unadjusted (but age-standardized) analyses of race/ethnicity differences in risks, Non-Hispanic Blacks and US-born Mexican Americans were more likely than Non-Hispanic Whites and Foreign-born Mexican Americans to be in the highest CHD risk categories, and both US-born and Foreign-born Mexican Americans had higher prevalence of metabolic syndrome / diabetes than Non-Hispanic Whites (data not shown). However, after stratification by gender and SES, race/ethnicity disparities in global CHD risk were not consistently seen in all strata. For instance, only middle-SES, Non-Hispanic Black women and high-SES, US-born Mexican American men had increased risk compared to the majority group at comparable SES (Table 4a); the largest disparity was for high-SES, US-born, Mexican American men: Compared to high-SES Non-Hispanic White men, their OR for being in the highest risk group was 4.1 (95% CI: 1.8 to 9.2). Foreign-born Mexican Americans (both men and women) showed reduced CHD risk at multiple SES levels, the largest advantage was in middle-SES, foreign-born Mexican American women: Compared to middle-SES Non-Hispanic White women, their OR for being in the highest risk group was 0.1 (95% CI: 0.01 to 0.4). These few race/ethnicity differences were not substantially affected by controls for lifestyle factors and central obesity – See Table 4a.

Table 4a
Ethnic Differences in Estimated 10-year CHD Risk (Odds Ratios Compared to Non-Hispanic Whites, with 95% Confidence Intervals), Stratified by Socioeconomic Status (SES), Before and After Adjusting for Physical Activity, Current Smoking, and Central Obesity; ...

Race/ethnicity disparities in metabolic syndrome/diabetes risk were also not uniform across gender/SES strata: High-SES, Non-Hispanic Black men, middle-SES, Non-Hispanic Black women, middle and high-SES, US-born Mexican American men, middle-SES, foreign-born Mexican American men, and low-SES, foreign-born Mexican American women had higher prevalence of diabetes than Non-Hispanic Whites of the same gender and SES (Table 4b); the largest disparity was for high-SES, US-born Mexican American men: Compared to high-SES, Non-Hispanic White men, their diabetes OR was 3.0 (95% CI 1.4 to 6.1). These race/ethnicity differences were not substantially affected by controls for lifestyle factors – See Table 4b.

Table 4b
Ethnic Differences in Metabolic Risk (Odds Ratios Compared to Non-Hispanic Whites, with 95% Confidence Intervals), Stratified by Socioeconomic Status (SES), Before and After Adjusting for Physical Activity and Current Smoking; United States 2001-2006


Our objective was to determine the independent associations of SES and race/ethnicity with global CHD risk and metabolic syndrome/diabetes prevalence in the US population. SES was strongly, inversely associated with global CHD risk in both men and women from all race/ethnicity groups, except in Foreign-born Mexican American men. SES was also strongly, inversely associated with metabolic risk in women from three of the four race/ethnicity groups and in Non-Hispanic White men. Stronger SES associations in women than in men have been previously reported (15,33,34,35), consistent with stronger SES-obesity relationships in women (36). Previous studies have also found that education and income gradients in health are weaker in African Americans than in Caucasian Americans (11,12,13,14,34); however, SES, defined by combining education and income, showed strong CHD risk gradients in all four race/ethnicity groups. This inconsistency is likely the result of differential employment and income returns from education (15,37,38,39), so that education or income on its own does not adequately capture social standing.

Some but not all of the SES associations with global CHD risk were explained by differences in health behaviors (physical activity and smoking) and central obesity. Other mediators of SES effects are likely to include neighborhood deprivation, access to medical care, and the physiological impact of daily stresses related to social position.

In contrast to the strong SES gradients with risks, race/ethnicity differences in risks were not consistent across SES and gender strata: Compared to Non-Hispanic Whites, global CHD risk was higher only in upper-SES, US-born, Mexican American men and middle-SES, Non-Hispanic Black women, and lower in foreign-born, Mexican American men and women. Diabetes prevalence was higher in minority groups, but only in some SES strata, and metabolic risk was actually lower in middle-SES, foreign-born Mexican American women. Previous studies that have examined ethnic disparities after stratifying by SES have also found black-white differences in risks only in middle and high SES strata, but these studies did not separate men from women, and did not examine predicted global CHD risk (9). Lower CHD risk in foreign-born Mexican Americans has been previously noted (29,30,31), and it has been recognized that the Hispanic paradox is actually an immigrant health phenomenon, resulting from selection pressures in migration that favor healthy immigrants (29,31,40), return emigration of sicker individuals (‘salmon bias’), and better lifestyle choices (41). Foreign-born Mexican Americans indeed had the lowest rates of current smoking and central obesity, but the risk reduction did not diminish with controls for lifestyle, suggesting that lack of acculturation is not a major explanation for this phenomenon.

The consistency of SES-risk gradients across gender and race/ethnicity strata, relative to the inconsistency of race/ethnicity disparities in risk across SES strata, suggest that SES, rather than race/ethnicity, is the main driver of social disparities in cardiovascular risk in the US. Previous studies in some populations have also found that SES eclipses race/ethnicity as a predictor of risks (42), and that SES explains much of the race/ethnicity disparities in health risks (43,44).

This study has some limitations that need to be acknowledged. The possibility of false discovery as a result of multiple testing hampers our ability to infer real race/ethnicity differences in risk, since differences were seen only in some, not all, SES strata; this is less of a concern with SES gradients, which were consistent across race/ethnicity groups and gender, but is a real possibility for ethnic differences that were detected, since they were seen in only some, not all, SES and gender strata.. Secondly, just as SES-risk associations varied by race/ethnicity, the effects of health behaviors on risks may also differ by ethnicity and SES (45), in which case, one cannot completely control for health behaviors by including them as covariates in the model. Also, abdominal obesity which was used as a surrogate for health behaviors related to diet and exercise, is also affected by family history and other factors, and so, control for health behaviors may not have been adequate. Finally, differential response by SES within race/ethnic groups could have biased our findings, although the use of weights designed to make the sample nationally representative should have diminished such bias.

The study’s limitations are outweighed by its strengths. In contrast to previous studies that have examined disparities in individual risk factors for CHD, we examined global risk for CHD. Disparities in individual risk factors can be in opposite directions (5,21) and effects of risk factors can vary by gender; thus disparities in a global risk measure are more relevant to actual outcomes. A few studies have previously examined disparities in global CHD risk, but have either studied only one gender (13) or not controlled for race/ethnicity (36) or SES (46,47). Since the predictive ability of the Framingham risk score may vary by SES and race/ethnicity (48), we also examined ORs for prevalent CHD or CHD risk-equivalent (highest risk group), and the disparities were similar to those for predicted risk. Secondly, unlike other studies of ethnic disparities, we conducted SES-stratified analyses, the more appropriate way to control for SES when SES associations differ between groups. Third, we separated US-born from foreign-born Mexican Americans, since their social experiences are distinct, and we found important differences.

In conclusion, this large, national study documents strong, inverse SES gradients with CHD risk in all race/ethnicity groups, and demonstrates that race/ethnicity disparities in risk are primarily due to SES differences between the groups. Healthy lifestyle choices and avoidance of central obesity appear to dampen SES associations. Socioeconomically disadvantaged individuals need to be specifically targeted for early risk detection and management and health behavior counseling if we are to improve the cardiovascular health of the nation.


The research was supported by the National Institutes of Health through two R01 grants, AG023347 (Crimmins, PI) and R01 AG026105 (Karlamangla, PI).


Coronary Heart Disease
Socio-economic Status
National Cholesterol Education Program
National Health and Nutrition Examination Survey
United States


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.


1. Williams DR, Collins C. US socioeconomic and racial differences in health: Patterns and explanations. Annu Rev Socio. 1995;21:349–86.
2. Marmot MG. Socio-economic factors in cardiovascular disease. J Hypertens Suppl. 1996;14(5):S201–S205. [PubMed]
3. Pamuk E, Makue D, Heck K, Reuben C, Lochner K. Socioeconomic status and Health Chartbook: Health, United States, 1998. National Center for Health Statistics; Hyatsville, MD: 1998.
4. Hunt KJ, Resendez RG, Williams K, Haffner SM, Stern MP, Hazuda HP. All-cause and cardiovascular mortality among Mexican-American and non-Hispanic white older participants in the San Antonio Heart Study – evidence against the “Hispanic paradox” Amer J Epidemiol. 2003;158:1048–1057. [PubMed]
5. Mensah GA, Mokdad AH, Ford ES, et al. State of disparities in cardiovascular health in the United States. Circulation. 2005;111:1233–1241. [PubMed]
6. Adler NE, Boyce WT, Chesney MA, Cohen S, Folkman S, Kahn RL, Syme L. Socioeconomic status and health: The challenge of the gradient. American Psychologist. 1994;49:15–24. [PubMed]
7. US Department of Health and Human Services Healthy People 2010. 2nd Edition With Understanding and Improving Health and Objectives for Improving Health. Washington DC: US Government Printing Office; Nov, 2000.
8. WHO World Health Report 2003: Shaping the Future. Geneva, Switzerland: WHO; 2003.
9. Sharma S, Malarcher AM, Giles WH, Myers G. Racial, ethnic and socioeconomic disparities in the clustering of cardiovascular disease risk factors. Ethn Dis. 2004;14(1):43–48. [PubMed]
10. Centers for Disease Control and Prevention Racial/ethnic and socioeconomic disparities in multiple risk factors for heart disease and stroke – United States, 2003. MMWR. 2005;54(05):113–117. [PubMed]
11. Jacobs DR, Jr., Burke GL, et al. Relationships of low density lipoprotein cholesterol with age and other factors: A cross-sectional analysis of the CARDIA study. Annals of Clinical Research. 1988;20(1-2):32–38. [PubMed]
12. Watkins LO, Neaton JD, Kuller LH. Racial differences in high-density lipoprotein cholesterol and coronary heart disease incidence in the usual-care group of the Multiple Risk Factor Intervention Trial. Am J Cardiol. 1986;57(8):538–545. [PubMed]
13. Kraus JF, Borhani, Franti CE. Socioeconomic status, ethnicity, and risk of coronary heart disease. Am J Epidemiol. 1980;111(4):407–414. [PubMed]
14. Karlamangla AS, Singer B, Williams D, Schwartz J, Matthews K, Kiefe C, Seeman TE. Impact of socioeconomic status on accumulation of cardiovascular risk factors in young adults: The CARDIA Study. Social Science and Medicine. 2005;60:999–1015. [PubMed]
15. Winkleby MA, Cubbin C, Ahn DK, Kraemer HC. Pathways by which SES and ethnicity influence cardiovascular risk factors. Annals NY Acad Sci. 1999;896:191–209. [PubMed]
16. Clark R, Anderson NB, Clark VR, Williams DR. Racism as a stressor for African Americans: A biopsychosocial model. Am Psychol. 1999;54:805–816. [PubMed]
17. Meyer IH. Prejudice as stress: Conceptual and measurement problems. Am J Pub Health. 2003;93(2):262–265. [PubMed]
18. Kessler RC, Neighbors HW. A new perspective on the relationships among race, social class, and psychological distress. J Health and Soc Behav. 1986;27:107–115. [PubMed]
19. Winkleby MA. Accelerating cardiovascular risk factor change in ethnic minority and low socioeconomic groups. Annals Epidemiol. 1997;S7:S99–S103.
20. Farmer MM, Ferraro KF. Are racial disparities in health conditional on socioeconomic status? Soc Sci Med. 2005;60:191–204. [PubMed]
21. Hertz RP, Unger AN, Ferrario CM. Diabetes, hypertension, and dyslipidemia in Mexican Americans and Non-Hispanic Whites. Am J Prev Med. 2006;30(2):103–110. [PubMed]
22. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) JAMA. 2001;285:2486–2497. [PubMed]
23. US Preventive Services Task Force Aspirin for the primary prevention of cardiovascular events: Recommendation and rationale. Ann Intern Med. 2002;136:157–160. [PubMed]
24. Pearson TA, Blair SN, Daniels SR, et al. AHA guidelines for primary prevention of cardiovascular disease and stroke: 2002 update. Consensus panel guide to comprehensive risk reduction for adult patients without coronary or other atherosclerotic vascular diseases. Circulation. 2002;106:388–391. [PubMed]
25. Grundy SM, Cleeman JI, Merz N Bairey, et al. for the NCEP Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III Guidelines. Circulation. 2004;110:227–239. [PubMed]
26. Balkau B, Charles MA, for the European Group for the Study of Insulin Resistance (EGIR) Comment on the provisional report from the WHO consultation. Diabetic Medicine. 1999;16:442–443. [PubMed]
27. Centers for Disease Control and Prevention (CDC) National Center for Health Statistics (NCHS) National Health and Nutrition Examination Survey Data. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; Hyattsville, MD: 2001-2006.
28. Center for Disease Control and Prevention. National Center for Health Statistics . National Health and Nutrition Examination Survey 2001-2002 Data Release. Sample Person Questionnaire: Demographics. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; Hyattsville, MD: May, 2006.
29. Palloni A, Arias E. Paradox lost: Explaining the Hispanic adult mortality advantage. Demography. 2004;41(3):385–415. [PubMed]
30. Crimmins EM, Kim JK, Alley DE, Karlamangla A, Seeman T. Hispanic paradox in biological risk profiles. Am J Public Health. 2007;97(7):1305–1310. [PubMed]
31. Huh J, Prause JA, Dooley CD. The impact of nativity on chronic diseases, self-rated health and comorbidity status of Asian and Hispanic immigrants. J Immigr Minor Health. 2008;10(2):103–118. [PubMed]
32. MacIntyre S, Hunt K. Socio-economic position, gender, and health: How do they interact? J. Health Psychology. 1997;2(3):315–334. [PubMed]
33. Thurston RC, Kubzansky LD, Kawachi I, Berkman LF. Is the association between socioeconomic position and coronary heart disease stronger in women than in men? Am J Epidemiol. 2005;162(1):57–65. [PubMed]
34. Borrell LN, Dallo FJ, White K. Education and diabetes in a racially and ethnically diverse population. Am J Public Health. 2006;96:1637–1642. [PubMed]
35. Loucks EB, Magnusson KT, Cook S, Rehkopf DH, Ford ES, Berkman LF. Socioeconomic position and the metabolic syndrome in early, middle, and late life: Evidence from NHANES 1999-2002. Ann Epidemiol. 2007;17:782–790. [PubMed]
36. Reynes JF, Lasater TM, Feldman H, Assaf AR, Carleton RA. Education and risk factors for coronary heart disease: Results from a New England community. Am J Prev Med. 1993;9(6):365–371. [PubMed]
37. Williams DR. Race/ethnicity and socioeconomic status: Measurement and methodological issues. International Journal of Health Services. 1996;26:483–505. [PubMed]
38. Williams DR, Collins C. Racial residential segregation: A fundamental cause of racial disparities in health. Public Health Rep. 2001;116:404–416. [PMC free article] [PubMed]
39. Kaufman JS, Cooper RS, McGee DL. Socioeconomic status and health in blacks and whites: The problem of residual confounding and the resiliency of race. Epidemiology. 1997;8:621–628. [PubMed]
40. Razum O, Zeeb H, Rohrmann S. The ‘healthy migrant effect’ – not merely a fallacy of inaccurate denominator figures. Intl J Epidemiol. 2000;29:191. [PubMed]
41. Franzini L, Ribble JC, Keddie AM. Understanding the Hispanic paradox. Ethnicity & Disease. 2001;11(3):496–518. [PubMed]
42. Scuteri A, Vuga M, Najjar SS, Mehta V, Everson-Rose SA, Sutton-Tyrell K, et al. Education eclipses ethnicity in predicting the development of the metabolic syndrome in different ethnic groups in midlife: The Study of Women’s Health Across the Nation (SWAN) Diabet Med. 2008;25:1390–1399. [PMC free article] [PubMed]
43. Hayward MD, Crimmins EM, Miles TP, Yang Y. The significance of socioeconomic status in explaining the racial gap in chronic health conditions. Amer Sociol Rev. 2000;65:910–930.
44. Matthews KA, Sowers MF, Derby CA, Stein E, Miracle-McMahill H, Crawford SL, Pasternak RC. Ethnic differences in cardiovascular risk factor burden among middle-aged women: Study of Women’s Health Across the Nation (SWAN) Am Heart J. 2005;149(6):1066–1073. [PubMed]
45. Paek K-W, Chun K-H, Jin K-N, Lee K-S. Do health behaviors moderate the effect of socioeconomic status on metabolic syndrome? Ann Epidemiol. 2006;16:756–762. [PubMed]
46. Ford ES, Giles WH, Mokdad AH. The distribution of 10-year risk for coronary heart disease among US adults. Findings from the National Health and Nutrition Examination Survey III. J Am Coll Cardiol. 2004;43:1791–1796. [PubMed]
47. Sundborn G, Metcalf PA, Gentles D, Scragg RKR, Schaaf D, Dyall L, et al. Ethnic differences in cardiovascular risk factors and diabetes status for Pacific ethnic groups and Europeans in the Diabetes Heart and Health Survey (DHAH) 2002-2003, Auckland New Zealand. NZMJ. 2008;121(1281):28–39. [PubMed]
48. Brindle PM, McConnachie A, Upton MN, Hart CL, Smith G Davey, Watt GC. The accuracy of the Framingham risk score in different socioeconomic groups: A prospective study. Br J Gen Pract. 2005;55(520):838–845. [PMC free article] [PubMed]