The racial-ethnic distribution of the United States continues to evolve with African Americans (AA) forming almost 13.4 of the population as per the U.S. Census Bureau. Further African Americans face a disproportionately higher prevalence of cardiovascular diseases as compared to other racial-ethnic groups. According to the Centers for Disease Control and Prevention (CDC), African Americans had the highest rates of deaths due to heart diseases (208 per 100,000) in 2018 as compared to other racial/ethnic groups (168.9 among Whites, 114.1 among Hispanics and 85.5 among Asian or Pacific Islanders respectively) . Long term public health efforts have led to a steady decline in deaths attributable to heart diseases (337.4 to 208/100,000) in 2000 to 208/100,000) in 2017; however, prevalence of risk factors of cardiometabolic diseases including hypertension, obesity, diabetes, and high cholesterol have increased . Recent National Center for Health Statistics from 2015-2016 identified prevalence of hypertension, obesity, diabetes and high cholesterol to be 42.1%, 47.5%, 19.6% and 10.2% respectively among US adults .
Metabolic syndrome (MetS) remains a significant clinical and public health burden in the US with nationwide estimates suggesting that 34% of US adults have MetS and prevalence among AA men and women is estimated to be 23% and 35% respectively  . MetS came into existence based on conceptual framework that argued the role of insulin resistance as a common mechanism underlying the associated abnormalities of blood pressure, HDL-cholesterol, triacylglycerol and glucose tolerance . MetS has been associated with an increase in overall mortality, both all cause as well as cardiovascular specific and cancer-specific mortality  . However, there are limitations which prevent the use of metabolic syndrome as a diagnostic or management tool due to its complex definitions and differential clustering of cardiometabolic risk factors   . To reduce redundancy, experts from several fields implemented a Joint Interim Statement of the International Diabetes Federation Task Force on Epidemiology and Prevention to develop a joint definition of MetS as described in methods . Yet, this definition does not identify clustering of factors that form basis of MetS that might make it clinically relevant and help develop targeted interventions.
The etiological basis of MetS remains multifactorial and complex. However, limited evidence exists on the impact of socioeconomic factors on risk of MetS and its components among African Americans. AA continues to face long-term impacts of exposure to adverse situations including but not limited to low income, racism and discrimination as well as racial and neighborhood segregation limiting their access to quality care . Studies among AA demonstrate higher risk of MetS among lower socioeconomic status (SES) as measured by either household income or education, however results remain controversial with studies reporting differential associations   .
To address these gaps, it is important to assess the impact of individual’s SES on prevalence of MetS and its components among the AA community. Further, given that MetS and chronic diseases impact men and women differentially, these associations need to be explored further across gender groups. Further, it becomes imperative to go beyond current landscape of MetS and assess clustering of cardiometabolic risk factors that define MetS among African Americans as this remains key for health promotion and disease prevention. Hence our study aims to address these gaps in literature on MetS and poverty among African Americans. This project aims to assess the impact of poverty status on prevalence of MetS and its components among AA men and women. Our secondary aim is to assess clustering of these components across AA men and women
2.1. Study Population Characteristics
The National Health and Nutrition Examination Survey (NHANES) is undertaken by the National Center for Health Statistics (NCHS) of the US Centers for Disease Control and Prevention (CDC) to assess the health and nutritional status of adults and children in the US. NHANES uses a multistage, stratified and clustered probability sampling strategy in which Mexican Americans, AA and the elderly are oversampled to ensure adequate sample size and to represent the total US civilian, non-institutionalized population. We applied survey weights to estimate US based generalized estimates . Since 1959, the NCHS has collected, analyzed and disseminated NHANES data that represents U.S. individuals from all 50 states. The NHANES includes an interview and an examination that includes anthropometric measurements. Data are available for public use. We utilized data from 3 NHAHES cycles: 2001-2002, 2003-2004, and 2005-2006. Details of the survey design, methods and data collections are available on the NHANES website (https://www.cdc.gov/nchs/NHANES/index.htm. Accessed 30 Jan. 2017). Our inclusion criteria 1) Participants of African American race; 2) 18 years and older; 3) Have available information on MetS and its components. Exclusion criteria included: 1) Women who were pregnant; 2) Missing information on MetS.
2.2. Definition of Metabolic Syndrome
We used the Joint Scientific Statement on metabolic syndrome to define MetS . The joint statement recently developed characterizes individuals on whether they have metabolic syndrome or not based on presence or absence of 3 out of 5 criteria’s: 1) elevated waist circumference (WC) (≥88 cm for women and ≥102 cm for men); 2) elevated triglycerides (TGL) (≥150 mg/dL) or drug treatment for elevated triglycerides; 3) low HDL cholesterol (HDL) (<40 mg/dL for men and <50 mg/dL for women) or drug treatment for low HDL cholesterol; 4) elevated blood pressure (BP) (systolic ≥130 mm Hg, or diastolic ≥85 mm Hg, or both) or antihypertensive drug treatment for a history of hypertension; and 5) elevated fasting glucose (FPG) (≥100 mg/dL) or drug treatment for elevated glucose. Details of collection procedures for these components have been described previously  .
2.3. Definition of Poverty Status
NHAHES provides each individual a unique value for their Poverty to income ratio (PIR). PIR is defined based on the US Census Bureau definition and calculated for all participants enrolled in NHANES is based on Department of Health and Human Services poverty guidelines that are issued each year in the Federal register for determining financial eligibility for federal programs . It is measured by dividing the family income by the poverty threshold income based on guidelines specific to family size and year. Values ranged from 0 to 5. We defined 3 levels of poverty status based on following cut offs: PIR < 1: below poverty, PIR of 1 - 3: above poverty & PIR of >3: high income based on prior reports.
2.4. Definition of Other Covariates
Age was assessed as a continuous variable. Education, based on self-report, was defined as some college or college graduate & higher, general education diploma (GED) or high school, and less than high school. Marital status was classified as married and not married (single, divorced or widowed). Participants were classified as never, former and current smokers from self-reported information. Participants were asked if they had smoked more than 100 cigarettes in their lifetime and whether they were smoking currently or not. If participants smoked >= 100 cigarettes in their lifetime and were currently smoking, they were classified as current smokers, if they smoked >= 100 cigarettes in their lifetime and had quit smoking, were classified as former smokers and those that smoked < 100 cigarettes in their lifetime were classified as never smokers. Alcohol consumption was defined based on participant’s self report of having 12 drinks in their lifetime. If participants reported yes, they were classified as drinkers and if they reported having consumed < 12 drinks, they were classified as non-drinkers. Vigorous physical activity information was obtained from the questions on whether participants did any activity that caused heavy sweating or large increases in breathing or heart rate (e.g., swimming, aerobics, or fast cycling), while moderate physical activity was determined from the questions on whether they did any activities that caused light sweating or a moderate increase in the heart rate (e.g., playing golf, dancing, bicycling for pleasure, or walking) . For analysis, participants who reported participating in either vigorous or moderate were coded as participating in any physical activity, whereas participants who reported “no” to both were coded as not participating in any physical activity. Diet quality was a continuous score based on the Healthy Eating Index (HEI) 2010 score. The Healthy Eating Index (HEI) is a measure of diet quality in terms of conformance with federal dietary guidance . The HEI-2010 includes 12 components, 9 of which assess adequacy of the diet by summing total intake gained from 1) total fruit; 2) whole fruit; 3) total vegetables; 4) greens and beans; 5) whole grains; 6) dairy; 7) total protein foods; 8) seafood and plant proteins; and 9) Fatty acids. The remaining 3 include refined grains, sodium intake and diet from empty calories (i.e., energy from solid fats, alcohol, and added sugars), assess dietary components that should be consumed in moderation . For all components, higher scores reflect better diet quality because the moderation components are scored such that lower intakes receive higher scores. The scores of the 12 components are summed to yield a total score, which ranges to 100 .
3. Statistical Analysis
Baseline population characteristics were reported as either continuous (mean and standard deviation) or categorical (N and %). We used sampling weights incorporated in NHAHES cycles that help account for account for selection probabilities, over-sampling, non-response, and differences between the sample and the total US population and help produce generalizable results. For our analysis, we performed survey weighted multivariable logistic regression models stratified by sex to study the association between poverty status and MetS across gender groups. These models were adjusting for sociodemographic characteristics including age, education status and marital status and lifestyle factors including smoking status, alcohol intake, physical activity and overall HEI 2010 score. In subgroup analysis, we performed weighted multivariable models were used to assess poverty status and the prevalence of individual metabolic syndrome components, stratified by sex. For these analyses, we used high income (PIR > 3) as the reference group. Two-sidedp value of <0.05 was statistically significant.
Finally, among participants with MetS, we assessed patterns of clustering of individual metabolic syndrome components among participants who had presence of 3 - 5 MetS components overall and by sex. These were presented using descriptive statistics i.e. number and % due to small sample sizes.
We identified 3233 African Americans in NHANES 2001-2006 that met our inclusion criteria. The mean age of participants was 43 (Standard Deviation: 0.42) years. Most of our participants were females (54%), had some college/college graduate (46%) and were more likely to be either single/widowed or divorced (56%) (Table 1). Approximately 33% were within high income group (PIR > 3), 43% lived above poverty (PIR 1 - 3) and 24% lived below poverty (PIR < 1). Majority of our participants had never smoked (53%), had consumed alcohol (60%) and participated in some type of physical activity (59%). Further, the mean score on HEI 2010 was 44.32 for our participants, lower than national average score of 56, indicating an overall poor diet .
Participants with MetS were more likely to be older, women, married, be ever/former smokers, and were less likely to be ever drinkers and participate in no physical activity (Supplemental TableS1 and TableS2). MetS participants were more likely to have higher scores on HEI 2010, even though average overall quality was poor.
4.1. Prevalence of MetS and MetS Components
The aggregate prevalence of MetS was 22%, with women having higher prevalence than men (25% vs 18%, p < 0.001). The prevalence of MetS components for men and women were as follows: elevated waist circumference (33% for men vs 70% for women, p < 0.0001), elevated triglycerides (34% for men vs 29% for women, p = 0.13); reduced HDL levels: (27% for men vs 38% for women, p < 0.001), elevated blood pressure (53% for men vs 51% for women, p = 0.17) and elevated plasma glucose (39% for men vs 38% for women, p = 0.57) (Figure 1(a)).
4.2. Association of Poverty Status with Metabolic Syndrome and Its Components
The results for the adjusted sex stratified association between poverty status, MetS and its components are presented in Table 2. For our analysis, we found women living below poverty status were more likely to have MetS than high income women (OR = 1.57, 95%CI = 1.00, 2.46, p = 0.05). In contrast we found that men living below poverty were less likely to have MetS as compared to high income men, though this was not statistically significant (OR = 0.70, 95%CI = 0.43, 1.19, p = 0.13).
Table 1. Demographic and lifestyle characteristics of African American participants enrolled in NHANES 2001-2006 (n = 3233).
Table 2. Multivariable Model to study the association of poverty status with metabolic syndrome and its components among African American men and women in NHAHES 2001-2006.
Multivariable Models adjusted for PIR, Age, Marital Status, Education, Smoking Status, Alcohol intake, Physical Activity and Healthy Eating Index.
Women living above poverty (PIR 1 - 3) and below poverty (PIR < 1) were more likely to have elevated waist circumference compared to high income women (OR = 1.70, 95%CI= 1.26, 2.30, p= 0.0001; OR = 2.04, 95%CI = 1.37, 3.01, p = 0.0001, respectively). Women living below poverty almost were twice as likely to have elevated triglycerides compared to high income women (OR = 1.85, 95%CI = 1.02, 3.36, p= 0.04). Women living below poverty was associated with two times the likelihood of having reduced HDL compared to high income women (OR = 2.04, 95%CI = 1.15, 3.60, p = 0.02). Women living above and below poverty over twice as likely to have elevated blood pressure compared to high income women (OR = 2.17, 95%CI = 1.60, 2.97, p = < 0.0001; OR = 2.16, 95%CI = 1.34, 3.49, p = 0.002, respectively). No association was observed between poverty status and fasting plasma glucose among women. Further, there was no association between poverty status and individual MetS components among men (Figure 1(b)).
(a) (b) (c)
Figure 1. (a) Prevalence of MetS components by Gender; (b) Odds Ratio for metabolic syndrome components by poverty status among AA men; (c) Odds Ratio for metabolic syndrome components by poverty status among women.
4.3. Clustering of Cardiometabolic Risk Factors
We identified 16 possible clusters of cardiometabolic risk factors that defined MetS using the joint definition (10 clusters of 3 factors, 5 clusters of 4 factors and 1 cluster of all 5 factors. Overall, the combination of WC + BP + FPG was most prevalent (17%), followed by clustering of WC + TGL + HDL + BP + FPG (17%) and WC + TGL + HDL + BP (16%) (Figure 2(a)). When stratified by gender, among AA men, clustering of TGL + HDL + BP was most common (17%), followed by WC + TGL + HDL + BP + FPG (16%) and WC + TGL + HDL + BP (4%). Among women, clustering of WC + BP + FPG (20%) was most
Figure 2. (a) Total No of participants with Mets = 738 Clustering of factors overall; (b) Clustering across sex.
common, followed by clustering of WC + TGL + HDL + BP (19%) and WC + TGL + HDL + BP + FPG (17%) (Figure2(b)). Though we further stratified to assess patterns of clustering across levels of poverty and gender, small sample sizes made it difficult to assess patterns or trends in distribution of each clustering type by levels of poverty and gender (Supplemental FigureS1(a) and FigureS1(b)). Overall, the contribution/prevalence of lipid levels (TG and HDL) were low among women but were highly prevalent among men highlighting need to identify and focus on these components while assessing MetS among AA men and women.
Our study is consistent with previous reports of poverty status and the risk of MetS among AA in the United States. Using a nationally representative data from the United States, we found that the aggregate prevalence of Met-S was higher among AA women. We also found that AA women had higher prevalence of elevated waist circumference and reduced HDL compared to AA men. In multivariable analysis, we found that compared to AA women living above poverty (i.e. PIR > 3), AA women living below poverty had higher odds of having metabolic syndrome, in contrast no increased odds observed among men. When considering the top 3 clusters, combination of TGL + HDL + BP was most prevalent among men, whereas among women combination of WC + BP was common across top 3 clusters.
According to the U.S. Census bureau, AA form approximately 13.4% of the U.S. population and are the largest racial-ethnic minority groups in the U.S. . African Americans have been subjected to systemic racism and discrimination that has had an impact on their overall health, financial and economic growth and impacted their health through several mechanisms . For example, AA bear disproportionately higher burden of morbidity, mortality, disability and injury as compared to Non-Hispanic Whites . AA face high rates of premature death due to cardiovascular disease, diabetes, hypertension and obesity as compared to other racial-ethnic groups . Multifactorial reasons exists underlying these health disparities: cultural practices for lifestyle patterns, inherited health risks, inadequate and lack of access to health care, variations in socioeconomic status, and residential segregation    . Another important factor involves the role chronic stressors and personalized experiences have on overall health outcomes, including poor self-rated health and high blood pressure .
Several previous studies have also reported that lower socioeconomic level is associated with increased risk of MetS among women but not men. Park et.al examined the association of poverty status on risk of metabolic syndrome among Korean individuals 20 to 79 years who participated in the Korean Health and Nutrition Examination survey . The authors found that as compared to women living below poverty, those living in the middle income (OR = 0.90, 95%CI = 0.75 - 1.08) and upper income (OR = 0.80, 95%CI = 0.66 - 0.97) groups had lower odds of MetS, and no associations were observed among men . Similar to our study, Loucks et al. also assessed socioeconomic factors and the risk of metabolic syndrome in NHANES III but used the National Institutes of Health, National Heart, Lung and Blood Institute Guidelines to define MetS . They found that among women and not men, low socioeconomic position as measured by education and PIR was associated with higher odds of metabolic syndrome and its components among participants enrolled in NHANES III . A study using data from the Atherosclerosis Risk in Communities Study found that individual-level and neighborhood-level SES factors were associated with risk of MetS among black and white women and not men .
One of the possible underlying mechanisms/models for the association between living below poverty level and increased prevalence of metabolic syndrome includes the reserve capacity model put forth by Gallo and Matthews in 2003 . The model suggests that people living in environments around low SES are prone to be exposed to harmful or difficult situations at higher frequency relative to individuals living in high SES regions, comparatively enjoy less benefits or rewarding situations while dealing with similar levels of stress . Further, low SES individuals have limited resources or reserve capacity (finances, peer support, access to health care) to deal with stress leading to increase in risk of chronic diseases . The reserve capacity model was tested among women enrolled in the Pittsburgh Healthy Women Study and authors found that low SES might be related directly to MetS or indirectly through low reserve capacity leading to heightened negative emotions and increase risk of MetS .
Differences in association of poverty with MetS for men and women may be explained by several factors. Women who pursue higher education generally have lower rates of birth. Pregnancy is associated with long term maternal effects including adiposity and decreased HDL levels so that women with higher education may be less likely to experience these MetS components  . Low income might be related to development of obesity. As such associated health risks could be viewed as a stigma among low socioeconomic women as compared to men that might possibly explain this association . Previous studies have shown that women are more sensitive to health inequalities beyond social stratification as compared to men, highlighting possible underlying mechanisms that need to be explored . Similarly, studies have also reported socioeconomic position including income is inversely associated with risk of several individual MetS components including waist girth among women  . Additionally, the likelihood of unemployment maybe greater in low income women leading to increase stress, low physical activity and poor choice of lifestyle factors . Men living below poverty might be more likely to have physically demanding professional activities that might increase their overall energy expenditure and thereby reduce their risk of MetS as observed in a previous study .
In our examination of the clustering of MetS components among men and women, we found that the clustering differed by sex. Though many underlying components of MetS share common pathophysiology that can explain future risk of diabetes and chronic diseases including cardiovascular diseases and cancer , MetS as a construct continues to create confusion due to its definition based on single components and hence it does not meet the definition of a single entity, leading to lack of consensus despite several efforts and lack of definitive treatment and management approaches . Hence understanding clustering of these factors can help provide clinicians with a definitive clinical plan to tackle risk of MetS as a whole and thereby reduce future risk of chronic diseases. Further, a previous study showed that the clustering of risk factors tend to remain constant from childhood to adulthood, and hence assessing clustering of factors across gender groups could help develop both population level and individual level health interventions to reduce overall burden of MetS distribution of these factors by racial-ethnic and gender groups could help adequately address burden of MetS .
Metabolic syndrome is an important premorbid condition and predictor of CVD and diabetes, however continues to face challenges including dichotomization of risk factors, attribution of relative as opposed to absolute risk, differing predictive value based on how definitions are used, inclusion of diabetics based on fasting plasma glucose and exclusion of other cardiometabolic risk factors for CVD . Though no clear understanding of pathophysiology of MetS exists, clustering can help identify mechanisms that increase risk of CVD. Some of these could include chronic activation of the immune system; induction of epigenetic or genetic changes based on participants lifecourse, disorders of the hypothalamic-pituitary-adrenal axis; altered glucocorticoid hormone action; chronic stress; and the contributions of cytokines, hormones and other molecules produced by adipocytes and further explain mechanisms that go beyond tradition CVD risk factors .
Our study has several limitations. The cross-sectional nature of the NHANES limits inference regarding causality due to lack of follow up data. Furthermore, we might have excluded other factors (not measured in NHANES) including genetic factors that might have influence on the risk of MetS. Neighborhood level SES factors are not available in publicly available NNHANES datasets. Hence, these should be explored in other future prospective cohort studies. The strengths of our study include the size generalizable nature of NHANES to the U.S. population. To reduce confusion caused by different definitions of MetS, we decided to use the Joint Scientific Statement of MetS. We explored clustering of factors to identify key components that can define MetS by sex among AAs.
Metabolic syndrome continues to remain an important predictor of cardiovascular disease, overall quality of life and risk of overall morbidity and mortality. Further steps towards defining specific clusters across levels of socio-economic factors to appropriately address the growing epidemic of MetS and develop appropriate population-specific public health interventions.
Innovation: Our work adds to the existing literature of the impact of socioeconomic characteristics on risk of cardiometabolic risk factors with focus on African Americans and further helps explore clustering of components that make up MetS among African American men and women. Further studies should consider long term impact of these clusters on risk of cardiovascular diseases and associated mortality.
AA women living below poverty are at higher risk of MetS and its components. Future studies should assess clustering of metabolic risk factors across age groups among AA men and women.
Ethics approval and consent to participate: Since this was a secondary data analysis of publicly available dataset, no ethics approval was needed.
Availability of data and materials: All data is publicly available on NHAHES website.
Table S1. Demographic and lifestyle characteristics of African American men participants enrolled in NHANES 2001-2006 (n = 1634) by PIR level.
Table S2. Demographic and lifestyle characteristics of African American women participants enrolled in NHANES 2001-2006 (n = 1599) by PIR level.
Figure S1. (a) Clustering among AA men by PIR; (b) Clustering among women by PIR.
 Moore, J.X., Chaudhary, N. and Akinyemiju, T. (2017) Peer Reviewed: Metabolic Syndrome Prevalence by Race/Ethnicity and Sex in the United States, National Health and Nutrition Examination Survey, 1988-2012. Preventing Chronic Disease, 14, E24.
 Aguilar, M., Bhuket, T., Torres, S., Liu, B. and Wong, R.J. (2015) Prevalence of the Metabolic Syndrome in the United States, 2003-2012. JAMA, 313, 1973-1974.
 Simmons, R., Alberti, K., Gale, E., Colagiuri, S., Tuomilehto, J., Qiao, Q., et al. (2010) The Metabolic Syndrome: Useful Concept or Clinical Tool? Report of a WHO Expert Consultation. Diabetologia, 53, 600-605.
 Guize, L., Thomas, F., Pannier, B., Bean, K., Jego, B. and Benetos, A. (2007) All-Cause Mortality Associated with Specific Combinations of the Metabolic Syndrome According to Recent Definitions. Diabetes Care, 30, 2381-2387.
 Dallmeier, D., Larson, M.G., Vasan, R.S., Keaney, J.F., Fontes, J.D., Meigs, J.B., et al. (2012) Metabolic Syndrome and Inflammatory Biomarkers: A Community-Based Cross-Sectional Study at the Framingham Heart Study. Diabetology & Metabolic Syndrome, 4, 28.
 Benjamin, E.J., Virani, S.S., Callaway, C.W., Chamberlain, A.M., Chang, A.R., Cheng, S., et al. (2018) Heart Disease and Stroke Statistics—2018 Update: A Report from the American Heart Association. Circulation, 137, e67-e492.
 Alberti, K., Eckel, R.H., Grundy, S.M., Zimmet, P.Z., Cleeman, J.I., Donato, K.A., et al. (2009) Harmonizing the Metabolic Syndrome: A Joint Interim Statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation, 120, 1640-1645.
 Noonan, A.S., Velasco-Mondragon, H.E. and Wagner, F.A. (2016) Improving the Health of African Americans in the USA: An Overdue Opportunity for Social Justice. Public Health Reviews, 37, Article No. 12.
 Park, M.J., Yun, K.E., Lee, G.E., Cho, H.J. and Park, H.S. (2007) A Cross-Sectional Study of Socioeconomic Status and the Metabolic Syndrome in Korean Adults. Annals of Epidemiology, 17, 320-326.
 Loucks, E.B., Rehkopf, D.H., Thurston, R.C. and Kawachi, I. (2007) Socioeconomic Disparities in Metabolic Syndrome Differ by Gender: Evidence from NHANES III. Annals of Epidemiology, 17, 19-26.
 Dallongeville, J., Cottel, D., Ferrières, J., Arveiler, D., Bingham, A., Ruidavets, J.B., et al. (2005) Household Income Is Associated with the Risk of Metabolic Syndrome in a Sex-Specific Manner. Diabetes Care, 28, 409-415.
 Moore, J.X., Chaudhary, N. and Akinyemiju, T. (2016) Metabolic Syndrome Prevalence by Race/Ethnicity and Sex in the United States, National Health and Nutrition Examination Survey, 1988-2012. Preventing Chronic Disease, 14, E24.
 Lopez, D.S., Advani, S., Tsilidis, K.K., Hernandez, M., Symanski, E., Strom, S.S., et al. (2014) Racial and Ethnic Differences in the Association of Metabolic Syndrome with Prostate-Specific Antigen Levels in US Men: NHANES 2001-2006. Journal of Men’s Health, 11, 163-170.
 Guenther, P.M., Casavale, K.O., Reedy, J., Kirkpatrick, S.I., Hiza, H.A., Kuczynski, K.J., et al. (2013) Update of the Healthy Eating Index: HEI-2010. Journal of the Academy of Nutrition and Dietetics, 113, 569-580.
 Guenther, P.M., Kirkpatrick, S.I., Reedy, J., Krebs-Smith, S.M., Buckman, D.W., Dodd, K.W., et al. (2014) The Healthy Eating Index-2010 Is a Valid and Reliable Measure of Diet Quality According to the 2010 Dietary Guidelines for Americans. The Journal of Nutrition, 144, 399-407.
 Mays, V.M., Cochran, S.D. and Barnes, N.W. (2007) Race, Race-Based Discrimination, and Health Outcomes among African Americans. Annual Review of Psychology, 58, 201-225.
 Fiscella, K. and Williams, D.R. (2004) Health Disparities Based on Socioeconomic Inequities: Implications for Urban Health Care. Academic Medicine, 79, 1139-1147.
 Krieger, N. (1991) Women and Social Class: A Methodological Study Comparing Individual, Household, and Census Measures as Predictors of Black/White Differences in Reproductive History. Journal of Epidemiology & Community Health, 45, 35-42. https://doi.org/10.1136/jech.45.1.35
 Krieger, N. (1999) Embodying Inequality: A Review of Concepts, Measures, and Methods for Studying Health Consequences of Discrimination. International Journal of Health Services, 29, 295-352.
 Krieger, N., Williams, D.R. and Moss, N.E. (1997) Measuring Social Class in US Public Health Research: Concepts, Methodologies, and Guidelines. Annual Review of Public Health, 18, 341-378.
 Krieger, N. and Sidney, S. (1996) Racial Discrimination and Blood Pressure: The CARDIA Study of Young Black and White Adults. American Journal of Public Health, 86, 1370-1378.
 Chichlowska, K.L., Rose, K.M., Diez-Roux, A.V., Golden, S.H., McNeill, A.M. and Heiss, G. (2008) Individual and Neighborhood Socioeconomic Status Characteristics and Prevalence of Metabolic Syndrome. The Atherosclerosis Risk in Communities (ARIC) Study. Psychosomatic Medicine, 70, 986.
 Gallo, L.C. and Matthews, K.A. (2003) Understanding the Association between Socioeconomic Status and Physical Health: Do Negative Emotions Play a Role? Psychological Bulletin, 129, 10.
 Matthews, K.A., Räikkönen, K., Gallo, L. and Kuller, L.H. (2008) Association between Socioeconomic Status and Metabolic Syndrome in Women: Testing the Reserve Capacity Model. Health Psychology, 27, 576.
 Finkelstein, E.A., Ruhm, C.J. and Kosa, K.M. (2005) Economic Causes and Consequences of Obesity. Annual Review of Public Health, 26, 239-257.
 Jeffery, R.W., Forster, J., Folsom, A., Luepker, R., Jacobs, J.D. and Blackburn, H. (1989) The Relationship between Social Status and Body Mass Index in the Minnesota Heart Health Program. International Journal of Obesity, 13, 59-67.
 Sarlio-Lähteenkorva, S. and Lahelma, E. (1999) The Association of Body Mass Index with Social and Economic Disadvantage in Women and Men. International Journal of Epidemiology, 28, 445-449.
 Santos, A.-C., Ebrahim, S. and Barros, H. (2007) Alcohol Intake, Smoking, Sleeping Hours, Physical Activity and the Metabolic Syndrome. Preventive Medicine, 44, 328-334.
 Magge, S.N., Goodman, E. and Armstrong, S.C. (2017) The Metabolic Syndrome in Children and Adolescents: Shifting the Focus to Cardiometabolic Risk Factor Clustering. Pediatrics, 140, e20171603.
 Camhi, S.M. and Katzmarzyk, P.T. (2010) Tracking of Cardiometabolic Risk Factor Clustering from Childhood to Adulthood. International Journal of Pediatric Obesity, 5, 122-129.