Social Determinants of Risk and Outcomes for Cardiovascular Disease: A Scientific Statement From the American Heart Association.

Abstract

HomeCirculationVol. 132, No. 9Social Determinants of Risk and Outcomes for Cardiovascular Disease Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBSocial Determinants of Risk and Outcomes for Cardiovascular DiseaseA Scientific Statement From the American Heart Association Edward P. Havranek, MD, FAHA, Mahasin S. Mujahid, PhD, MS, Donald A. Barr, MD, PhD, Irene V. Blair, PhD, Meryl S. Cohen, MD, FAHA, Salvador Cruz-Flores, MD, FAHA, George Davey-Smith, MA(Oxon), MD, BChir(Cantab), MSc(Lond), Cheryl R. Dennison-Himmelfarb, RN, PhD, FAHA, Michael S. Lauer, MD, FAHA, Debra W. Lockwood, Milagros Rosal, PhD and Clyde W. Yancy, MD, FAHAon behalf of the American Heart Association Council on Quality of Care and Outcomes Research, Council on Epidemiology and Prevention, Council on Cardiovascular and Stroke Nursing, Council on Lifestyle and Cardiometabolic Health, and Stroke Council Edward P. HavranekEdward P. Havranek Search for more papers by this author , Mahasin S. MujahidMahasin S. Mujahid Search for more papers by this author , Donald A. BarrDonald A. Barr Search for more papers by this author , Irene V. BlairIrene V. Blair Search for more papers by this author , Meryl S. CohenMeryl S. Cohen Search for more papers by this author , Salvador Cruz-FloresSalvador Cruz-Flores Search for more papers by this author , George Davey-SmithGeorge Davey-Smith Search for more papers by this author , Cheryl R. Dennison-HimmelfarbCheryl R. Dennison-Himmelfarb Search for more papers by this author , Michael S. LauerMichael S. Lauer Search for more papers by this author , Debra W. LockwoodDebra W. Lockwood Search for more papers by this author , Milagros RosalMilagros Rosal Search for more papers by this author and Clyde W. YancyClyde W. Yancy Search for more papers by this author and on behalf of the American Heart Association Council on Quality of Care and Outcomes Research, Council on Epidemiology and Prevention, Council on Cardiovascular and Stroke Nursing, Council on Lifestyle and Cardiometabolic Health, and Stroke Council Originally published3 Aug 2015https://doi.org/10.1161/CIR.0000000000000228Circulation. 2015;132:873–898Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2015: Previous Version 1 IntroductionAn Institute of Medicine report titled U.S. Health in International Perspective: Shorter Lives, Poorer Health documents the decline in the health status of Americans relative to people in other high-income countries, concluding that “Americans are dying and suffering from illness and injury at rates that are demonstrably unnecessary.”1 The report blames many factors, “adverse economic and social conditions” among them. In an editorial in Science discussing the findings of the Institute of Medicine report, Bayer et al2 call for a national commission on health “to address the social causes that have put the USA last among comparable nations.”Although mortality from cardiovascular disease (CVD) in the United States has been on a linear decline since the 1970s, the burden remains high. It accounted for 31.9% of deaths in 2010.3There is general agreement that the decline is the result, in equal measure, of advances in prevention and advances in treatment. These advances in turn rest on dramatic successes in efforts to understand the biology of CVD that began in the late 1940s.4,5 It has been assumed that the steady downward trend in mortality will continue into the future as further breakthroughs in biological science lead to further advances in prevention and treatment. This view of the future may not be warranted.The prevalence of CVD in the United States is expected to rise 10% between 2010 and 2030.6 This change in the trajectory of cardiovascular burden is the result not only of an aging population but also of a dramatic rise over the past 25 years in obesity and the hypertension, diabetes mellitus, and physical inactivity that accompany weight gain. Although there is no consensus on the precise causes of the obesity epidemic, a dramatic change in the underlying biology of Americans is not postulated. More likely culprits are changes in societal and environmental conditions that have led to changes in diet and physical activity. At the same time, there is increasing awareness that the benefits of advances in prevention and treatment have not been shared equally across economic, racial, and ethnic groups in the United States. Overall population health cannot improve if parts of the population do not benefit from improvements in prevention and treatment.The purpose of this statement is to increase awareness of the influence of social factors on the incidence, treatment, and outcomes of CVD; to summarize the current state of knowledge about these factors; and to suggest future directions in research, particularly research on effective interventions to attenuate or eliminate these adverse social influences. The statement is not intended to be a comprehensive review; references are intended to be illustrative and to highlight significant knowledge in the field. The premise underlying this scientific statement is that, at present, the most significant opportunities for reducing death and disability from CVD in the United States lie with addressing the social determinants of cardiovascular outcomes. Although social determinants are most often invoked in discussions of inequalities or disparities in health, we take a broader view that social factors can and do affect cardiovascular health in all. Thus, a consideration of the role of social determinants is essential if we are to achieve the American Heart Association 2020 Impact Goals: to improve cardiovascular health of all Americans by 20% while reducing deaths from CVD and stroke by 20%.7Defining Social Determinants of HealthThe World Health Organization defines the social determinants of health quite broadly as “the circumstances in which people are born, grow, live, work, and age, and the systems put in place to deal with illness.”8 This definition encompasses the view that health and illness are not distributed randomly throughout human society, and neither are resources to prevent illness and its effects. Instead, they cluster at the intersections of social, economic, environmental, and interpersonal forces.Cataloging the Social Determinants of HealthSocial determinants are highly interrelated and therefore difficult to catalog. Given our focus on CVD in the United States, this statement considers socioeconomic position (SEP; encompassing wealth and income, education, employment/occupational status, and other factors), race and ethnicity, social support (including social networks), culture (including language), access to medical care, and residential environments (Table 1). We additionally consider the psychological, behavioral, and biological mechanisms through which social determinants precipitate and perpetuate CVD.Table 1. Social Determinants of HealthSEPRace, ethnicitySocial supportCulture and languageAccess to careResidential environmentSEP indicates socioeconomic position.SEP and CVDDefining SEPThere are several ways to describe and measure social and economic conditions. The terms social class, social stratification, and social or socioeconomic status are used frequently and often interchangeably. Here, we use SEP, defined as the socially derived economic factors that influence what positions individuals or groups hold within the stratified structure of a society.9 Relations among groups within a society are determined largely by material circumstances, which in turn are determined by the relations these groups have with systems of economic production. Members of advantaged groups control resources (whether material, economic, political, social, or cultural) in a way that may exclude and dominate the disadvantaged. Unequal distribution and control over resources influence patterns of exposures, which act at different stages of the life course, resulting in unequal distribution of disease in different groups within a society. Health and SEP are seen as inextricably linked, with health itself seen as a marker of SEP in some schemes (Table 2). It is important to highlight that, although measured at the individual level, SEP is determined at least partly by structural relations between groups within society. For example, the level of education attained by an individual may be constrained by educational opportunities available to a particular group. We discuss area-level factors in more detail in the Residential Environments section.Measuring SEPThere is no single best indicator of SEP. Each indicator of SEP emphasizes a particular aspect of social stratification, which may be more or less relevant to different health outcomes and at different stages in the life course.9 Individual-level indicators of SEP include income, education, and occupation-based indicators, and ideally, they should be considered simultaneously. Others10 have emphasized that SEP should consider both actual resources and status as determined by prestige or rank-related characteristics. This multidimensional nature of SEP has been emphasized in the work of the Commission on the Measurement of Economic Performance and Social Progress11 headed by the Nobel Laureate Joseph Stiglitz. Although the Commission was focused on a critique of gross domestic product as a measure of the performance of societies or nations, some of its conclusions can be applied to individuals. Broadly, the commission concluded that well-being is determined by a number of interwoven dimensions (Table 2).Table 2. Markers of SEPMaterial conditions (based on income and wealth)HealthEducationAccess to valued personal activities (eg, work)Political voiceSocial connectionsEnvironmentPhysical insecurity (crime, violence)SEP indicates socioeconomic position.Associations With CVDThree measures of SEP have been explored extensively with regard to their relationship to cardiovascular health: education, income, and occupation. Broadly speaking, lower SEP in the United States is associated with a greater prevalence of CVD risk factors and a higher incidence of and mortality resulting from CVD. We highlight evidence linking measures of SEP with cardiovascular health, including early childhood socioeconomic conditions. Detailed and comprehensive reviews on this topic can be found elsewhere.12–14EducationEducation, the most used indicator of SEP in the United States, provides the most consistent results in relation to CVD outcomes.14,15 Lower levels of educational attainment are associated with a higher prevalence of cardiovascular risk factors (discussed in more detail in the Mechanisms Mediating the Relationship Between Societal Conditions and CVD section), higher incidence of cardiovascular events, and higher cardiovascular mortality, independent of sociodemographic factors.14,16 In relation to CVD mortality, Mackenbach et al17 examined educational differences in ischemic heart disease, cerebrovascular disease, and total CVD mortality in the United States and 11 Western European countries. They found higher mortality among individuals with lower education in all countries; however, the relative and absolute magnitude varied across countries.Disparities associated with educational attainment have widened over time. Meara et al18 used US Census and death certificate data to show that the disparity in life expectancy at 25 years of age between those with low educational attainment (≤12 years) and those with high educational attainment widened by 0.9 years from the 1980s to the 1990s. A widening education-based difference in cardiovascular death was responsible for 17.4% of the overall gap in life expectancy, second only to cancer. Similar increases in education disparities in life expectancy were documented between 1996 and 2006.19Low health literacy and numeracy might in part mediate the relationship between lesser education and CVD, with low health literacy associated with having less than a high school education and with poor health outcomes.20 Interventions that improve self-care behavior, risk factor control, or cardiovascular outcomes in those with low health literacy or numeracy are generally lacking. More study has been reported in heart failure, for which interventions have generally been resource intensive and results have been mixed.21Income and Income InequalityBoth income and income inequality have been studied in relation to cardiovascular health. Other measures of material circumstances beyond income, particularly accumulated wealth, have not been adequately considered in the literature. Findings for associations between income and cardiovascular health parallel those for education, with the caveat that many studies document nonlinearity in the association of income and cardiovascular outcomes.14 Data on >500 000 men and women from the National Longitudinal Mortality Study showed similar associations between education and income in relation to all-cause and cardiovascular mortality. After adjustment for sociodemographic factors, there was a 40% to 50% decrease in mortality with increasing levels of family income.19 Whether differences in cardiovascular outcomes are becoming more or less pronounced over time is unclear because income is more unstable and difficult to measure than education.22Income inequality within societies has grown in recent decades, particularly in high-income countries such as the United States, and the social consequence of this reality is becoming an important political issue. Harper et al22 found no evidence of consistent associations between income inequality and cardiovascular health, including prevalence of CVD risk factors and CVD trends.Employment/Occupational StatusThe relationship between occupation and CVD is less clear than it is for education or income. The Whitehall study was most influential in the description of differences in cardiovascular mortality by job classification. In the first cohort initiated in the 1970s, Marmot and colleagues23 followed up a group of 17 530 British civil servants in London, all of whom had office-based jobs and none of whom were considered to be economically disadvantaged. At the 10-year follow-up, mortality resulting from coronary heart disease was 2.2%, 3.6%, 4.9%, and 6.7% across job grade from the highest to the lowest; these differences remained significant after controlling for age, height, smoking, systolic blood pressure, cholesterol, and blood sugar. Comparable results have been documented with 25 years of follow-up.24 In the United States, Leigh and Du25 used data from the Health and Retirement Study to assess for an independent relationship between lifetime occupations grouped into 15 categories and prevalent hypertension, concluding that “in general, higher status occupations were associated with less hypertension.” Another study using National Health and Nutrition Examination Survey (NHANES) data26 tended to support this finding, additionally noting that protective service workers such as police and firefighters had the lowest rates of treatment for established hypertension. In general, there is a paucity of data on the relationship between occupation and cardiovascular morbidity and mortality in the United States.In addition to these relationships between type of employment and CVD, a relationship between unemployment and CVD has been postulated. Epidemiological studies of unemployment and health are particularly difficult because of potential “effect-cause” relationships, in which unemployment is a consequence of poor health rather than the reverse, and because of confounding by factors such as low educational attainment that might predict both unemployment and poor health. Nonetheless, the preponderance of evidence supports the position that job loss leads to illness. Studies of widespread labor downsizing, in which an individuals’ health is not a factor, support the causal relationship between job loss and ill health.27 At least for behavioral health issues, long-term longitudinal studies that gather health information before job loss have also supported a causal relationship.28 Studies specifically in CVD have been performed. Dupre and colleagues29 used data from the prospective Health and Retirement Study to study the relationship between unemployment and incident myocardial infarction. After adjustment for risk factors and sociodemographics, the hazard ratio for myocardial infarction was highest in the first year of unemployment and increased with the number of job losses.The possible psychological and biological mechanisms responsible for the relationship between occupation, unemployment, and CVD are discussed in the Mechanisms Mediating the Relationship Between Societal Conditions and CVD section.Life-Course Context of SEPFor CVD, poor socioeconomic conditions in early life appear to make an important contribution to disease risk in adulthood, especially when early-life factors influence the developmental trajectories of important adult risk factors.22 A systematic review of 40 studies investigating associations between childhood socioeconomic circumstances and ischemic heart disease, stroke, or combined CVD in adulthood reported that the majority of studies show robust associations of poorer childhood circumstances and CVD, although findings differed across types of CVD, socioeconomic measures, and sex.12 Galobardes et al9 reported heterogeneity in the strength of association of SEP indicators with specific CVDs, which suggests specificity of the pathogenic links between socioeconomically patterned exposures early in life and adult disease outcomes.22 Perhaps more important than the direct effect of early-life socioeconomic factors on CVD is their potential effect on the development of conventional risk factors.22 Reviews of studies focusing on the role of childhood socioeconomic conditions, usually indicated by the occupation or education of the parents, have found consistent evidence of an association with CVD risk factors such as blood pressure, lipid levels, body mass index (BMI), fibrinogen, smoking, physical activity, and alcohol consumption.22 Investigation of the impact of social mobility on social class inequalities in all-cause mortality has suggested a cumulative effect of lifetime socioeconomic experience. However, evidence that upward or downward socioeconomic mobility may play an important role in generating or substantially magnifying CVD differences is limited.22Further evidence suggests that the effect of early life socioeconomic conditions may depend on interactions with other risk factors in later life.22 Secular changes in CVD differentials are more congruent with increasing socioeconomic differences in cigarette smoking and consumption of micronutrients than with trends in socioeconomic differentials in infant mortality or height,30,31 understood as potential markers of early-life circumstances and related outcomes such as fetal growth. Thus, the interactions among early-life socioeconomic environments and risk factor trajectories seem to influence the development and maintenance of health behaviors and their cumulative biological sequelae as a major life-course process linking early-life SEP to CVD.22 The Mechanisms Mediating the Relationship Between Societal Conditions and CVD section provides a detailed discussion.SEP and CVD Risk PredictionGiven the substantial evidence linking SEP and CVD and findings that suggest that the Framingham risk score overestimates the risk of coronary heart disease in high-SEP individuals and underestimates the risk in low–socioeconomic status individuals, recent studies have begun to evaluate the potential benefit of including SEP in risk prediction models.32 Using data from the Atherosclerosis Risk in Communities (ARIC) Study and NHANES linked to the National Death Index, Fiscella and colleagues33 documented improvements in the calibration and reduction of bias in the Framingham risk model. These types of investigations should continue in future research.Recommendations and Conclusions: SEPNo single parameter fully captures SEP; income, education, and occupation have been used successfully.SEP measures may vary by race/ethnic groups, and these synergistic effects should be considered.Novel markers of SEP should be investigated for broader use in understanding CVD.Race/Ethnicity, Racism, and CVDFor this statement, we use the terms race and ethnicity as constructs with very little biological or genetic basis but as constructs shaped by the social, economic, and political forces of societies.4,5 Differences in health by race and ethnicity are a major public health concern.1 On the basis of projections from the US Census Bureau, the population of non-Hispanic whites will almost double by the year 2050, with Asians and Hispanics the fastest-growing populations in the United States. Racial and ethnic minorities are disproportionately burdened with poor health across a variety of different outcomes, and given the significant increase in these populations in the United States, attention to the health needs of these groups is essential.Racial/ethnic differences in cardiovascular health have been documented extensively.16 For example, in the Eight Americas Study, the authors identified 7 distinct groups within the United States based on race, geographic location, and income and found significant differences in life expectancy.34 These groups are, in order of decreasing life expectancy, Asian Americans (84.9 years), whites living in rural Northern Plains/Dakotas, low-income whites in Appalachia and the Mississippi Valley, western Native Americans, middle-income blacks, southern rural blacks, and blacks in poor urban areas (71.1 years), representing a 14-year difference between the highest and lowest group. CVD was the greatest source of differences in life expectancy. On the basis of the latest report of the American Heart Association heart and stroke disease statistics, blacks are 2 to 3 times more likely to die of heart disease compared with whites, and blacks and other racial/ethnic minorities have higher rates of premature death resulting from CVD and higher CVD risk factors.3 Declines in CVD mortality have not eliminated racial and ethnic differences in CVD; they remain constant.Although public opinion polls show that levels of overt or explicit racism have declined over the past 5 decades, there are clear indicators that members of ethnic minority groups, particularly blacks, must endure everyday slights and offenses that undermine health. Specific to CVD, studies have investigated the links between self-reported experiences of racism and both blood pressure and cardiovascular reactivity. The evidence to date shows limited direct relationships between reported racism and hypertension diagnosis or resting blood pressure measures.35 There is much stronger evidence, however, for ambulatory blood pressure monitoring, with all 6 known studies finding a positive relation between ambulatory blood pressure (particularly at night) and reports of racism or discrimination.36 In the largest of these studies,37 357 black and Latino adults completed a measure of lifetime experiences of ethnic or racial discrimination and then wore an ambulatory blood pressure monitor until they returned the next day. Both nighttime systolic blood pressure and diastolic blood pressure were positively related to amount of reported racism, even after adjustment for patient demographics and self-reported general hostility. Higher levels of reported racism were also associated with a lower likelihood of nocturnal dipping (≥10% decrease in nighttime blood pressure). Additional research has found that past experiences of racism predict greater cardiovascular reactivity.38 In 1 study, 165 black and white normotensive adults had their heart rates and blood pressures measured while they recalled an event that had made them angry. Participants who had earlier reported more experiences with discrimination were found to have greater heart rate and diastolic blood pressure reactivity during the recall task and slower recovery after the task, particularly if they were black and had a generally positive outlook on life (eg, low in cynicism or high in optimism).Of great concern to society is the possibility that healthcare provider bias contributes to the problem.39–43 Investigations of clinicians’ ethnic and racial attitudes have shown that, similar to the general population, clinicians show little explicit or intentional bias but exhibit substantial bias in their implicit (unconscious) attitudes.44–46 Theoretical models suggest that clinicians’ implicit bias may affect their delivery of health care in 3 ways.42,47,48 First, implicit bias may directly influence clinicians’ decisions about their patients’ medical treatment, with incorrect, often stereotypical assumptions leading to lower-quality care for minority than for white patients. A study by Schulman and colleagues40 used scripted videotaped interviews of actors portraying patients with chest pain, finding that physicians were less likely to recommend catheterization for black women than for white men reporting the same symptoms. The authors found no difference in the rate of physician-recommended catheterization for black men and white men. Green and colleagues45 found that resident clinicians with greater implicit bias were less likely to recommend thrombolytic therapy for a hypothetical black patient with myocardial infarction, but this did not occur when the patient was described as white. On the other hand, research on pediatric decision making49,50 showed that some hypothetical decisions were associated with implicit bias but others were not. However, a study51 with medical students failed to find any relation between clinical decisions in the hypothetical scenarios and the students’ implicit bias. Although this work is often criticized on methodological grounds, an influential review of the literature by the Institute of Medicine43 concluded that “bias, stereotyping, prejudice, and clinical uncertainty on the part of healthcare providers” may play a role in racial/ethnic health disparities. Thus, although proof of bias is difficult to achieve, it remains viable and of great concern.The second route by which implicit bias may affect care processes is by producing lower-quality clinical interactions and communication between (more biased) clinicians and minority patients. Several studies52–54 have found associations between clinicians’ implicit bias and worse clinical interactions with black patients. Most relevant to CVD is a study by Blair and colleagues44 in which primary care providers’ levels of implicit race bias predicted differences between black and white patients’ reports of their clinicians’ patient centeredness, with black patients reporting less patient centeredness for clinicians previously categorized as having higher levels of implicit racial bias. Numerous studies have investigated patients’ perceptions of bias and discrimination while receiving health care. A review of this literature55 found that up to 52% of blacks, 13% of Latinos, and 6% of non-Hispanic whites have reported biased treatment based on their race or ethnicity. Perceptions of biased treatment in turn have been associated with reports of lower health, lower levels of self-care or adherence, interruptions in care, mistrust of clinicians, and underuse of available services, although some studies have not found these associations.55 LaVeist and colleagues56 surveyed 781 black and 1003 white patients with serious chronic heart disease about their level of satisfaction with the care they received, their perceptions of trust in the healthcare system, and their perceptions of racial bias inherent in the healthcare system. In a multivariate analysis controlling for a range of demographic factors, they found a significant link between perceived racial bias in care, trust in the system, and satisfaction with care, with perceived bias predicting both lower trust and lower satisfaction.The third means through which race could have an adverse effect on medical care is stereotype threat.57 Stereotype threat occurs when individuals, often unconsciously, fear being judged negatively according to racial stereotypes. In the context of medical care, stereotype threat might cause a black patient to approach an ambulatory care visit concerned that he or she may be treated according to a stereotype such as being nonadherent with medications or less able to understand complex medical issues. The effect of stereotype threat on clinical interactions has seen limited study. In 1 report,58 an intervention known to blunt the effects of stereotype threat was administered to black patients about to see a primary care physician for hypertension care. Compared with those receiving a control intervention, those in the intervention group had patient-provider communication that was more interested, friendly, responsive, interactive, and respectful and was less depressed and distressed in tone.Recommendations and Conclusions: Race/Ethnicity, Racism, and CVDRace/ethnicity is a social construct with little biological or genetic basis.The concepts of implicit bias and stereotype threat are real phenomena that affect health and disease and may be root causes of disparate care.Effective interventions to improve patient-provider communication and patient satisfaction/trust across racial lines are clearly needed.Social Support, Social Networks, and CVDSocial SupportThe term social support has been defined in the literature