Inequalities and Deteriorations in Cardiovascular Health in Premenopausal US Women, 1990–2016

Abstract

Coronary heart disease (CHD) mortality rates in the United States have declined by up to two thirds in recent decades. Closer examination of these trends reveals substantial inequities in the distribution of mortality benefits. It is worrying that the uneven distribution of CHD that exists from lowest to highest social class—the social gradient—has become more pronounced in the United States since 1990 and is most pronounced for women.

Here we consider ways in which this trend disproportionately affects premenopausal women aged 35 to 54 years. We apply a social determinants of health framework focusing on intersecting axes of inequalities—notably gender, class, ethnicity, geographical location, access to wealth, and class—among other power relations to which young and middle-aged women are especially vulnerable, and we argue that increasing inequalities may be driving these unprecedented deteriorations. We conclude by discussing interventions and policies to target and alleviate inequality axes that have potential to promote greater equity in the distribution of CHD mortality and morbidity gains.

The application of this framework in the context of women’s cardiovascular health can help shed light regarding why we are seeing persistently poorer outcomes for premenopausal US women.

Over the past 50 years, a steady decline in deaths attributable to cardiovascular disease (CVD) has been observed in many industrialized countries. CVD mortality rates in the United States declined almost two thirds by 2000, with half of the decline in deaths from coronary heart disease (CHD) attributed to rapid progress in evidence-based treatments and risk factor reduction.¹ However, improvements in mortality have not been afforded to all people in the United States. Between 1980 and 1989, the estimated annual percentage change in CHD mortality was a sizeable reduction of −5.4% in premenopausal US women aged 35 to 54 years.¹ This change in CHD mortality dropped to a modest decrease of −1.2% from 1989 to 2000, and then increased to 1.5% from 2000 to 2002. Comparatively, men of this age group experienced a continual (albeit decelerated) decline in CHD mortality, with reductions of −6.2% (1980–1989), −2.3% (1989–2000), and −0.5 (2000–2002). Stagnation in young women’s CHD mortality declines has also been observed through 2011² and 2014.³ Over 20 years (1995–2014), there was an increase in incident myocardial infarction admission in US patients³—of note, in women aged 35 to 54 years. Overall, incident myocardial infarction admission rose among young patients from 27% (1995–1999) to 32% (2010–2014) but was most pronounced for young women. In fact, compared with young male patients, for whom admissions declined, women aged 35 to 54 years, particularly Black women, experienced a steady increase.³

Increases in CHD mortality and morbidity rates are largely unprecedented in industrialized societies. The average age of first myocardial infarction is 72 years for women compared with 65 years for men, yet this advantage may be negated when accounting for medical or lifestyle risk factors and when considering other types of CVD (e.g., cerebrovascular disease). This, coupled with data that acute coronary syndrome mortality rates are highest among women aged younger than 40 years,⁴ suggests there may be contemporary drivers of CVD risk for premenopausal women in the United States.

To date, much of the discussion around the worsening cardiovascular health of premenopausal US women has centered on temporal changes in individual-level markers of cardiovascular health. Notwithstanding the evidence that individual-level risk factors, particularly blood pressure, diabetes, alcohol use, smoking, and physical inactivity, are the major drivers of CVD in women, the burden of cardiovascular risk factors and mortality rates does not afflict all parts of the population equally. Disaggregation of population-level data reveal marked disparities in CVD deaths not only by sex or gender but also by other intersecting inequality axes including, but not limited to, social class, race/ethnicity, geographic location, disability, sexuality, religion, and access to wealth. The steep social gradient that has been observed in the context of CVD prevention and control⁵ requires consideration of the contemporary health challenges facing young, particularly non-White, women in the United States. In 2015, the American Heart Association released a position paper providing an overview of the social determinants of CVD risk and outcomes.⁶

Social determinants of cardiovascular health, by definition, tend to focus predominantly on socioeconomic position at the expense of other power relations that perpetuate inequalities. An intersectional framework recognizes that each does not occur in isolation—rather, they are intertwined.⁷ This article addresses such limitations by applying a social determinants of health framework that focuses on inequality axes including gender, race, and class—among other power relations—to which women can be especially vulnerable. We argue that increasing inequalities could be driving the unprecedented deterioration in US women’s cardiovascular health. We conclude by discussing interventions and policies that may have the potential to promote a more equitable distribution of CVD mortality and morbidity gains. The application of this framework can help shed light on the persistently poorer cardiovascular health outcomes for US women.

SOCIAL GRADIENT OF CARDIOVASCULAR DISEASE

The social gradient of CVD was first observed by Marmot et al. in the Whitehall II Study of British Civil Servants,⁸ in which they identified a pronounced inverse association between CVD mortality and class. In the years since, this social gradient has been replicated across different countries for a wide range of diseases. Indeed, the social gradient has been observed for CVD mortality in the United States. Singh and Siahpush ⁹ observed declining CVD mortality from 1969 to 1998 across all area-level socioeconomic groups with significantly greater mortality declines in higher socioeconomic groups. Using data from the National Longitudinal Mortality Study, the authors found that, in 1969, working-age persons in the United States who resided in the most deprived areas had 29%, 35%, and 73% higher heart disease, CVD, and stroke mortality, respectively.¹⁰ While absolute inequalities in CVD and heart disease mortality remained stable from 1969 to 2011, relative inequalities in mortality consistently widened owing to increasing socioeconomic disparities. This resulted in 120% higher mortality between 2007 and 2011. Socioeconomic gradients were observed both at the area level (e.g., unemployment rate, poverty rate, and housing quality) and individual level (e.g., education, income, and occupation) and were found to be steeper between 1990 and 2002 compared with between 1979 and 1989. For people in the United States aged 24 to 35 years, recent evidence indicates that disparities in individual-level education attainment are associated with poor cardiovascular health markers.¹¹

INEQUALITY AXES AND THE SOCIAL GRADIENT

These trends demonstrate that US women are not equally susceptible to the social gradient of CVD. Cardiovascular health is determined by the political and socioeconomic context (structural determinants) and driven by intersecting axes of inequality to shape material resources like housing and working conditions (intermediate determinants). Inequality axes relate to power relations (e.g., gender, ethnicity, geographic region, wealth; Figure 1) that intersect to have an impact upon an individual’s cardiovascular risk. These upstream and intermediate determinants, in turn, influence psychological and biobehavioral factors (downstream determinants) in a bidirectional manner including one’s interaction with the health care system. These determinants can exert unique influences at specific stages of the life span and accumulate to induce disease.¹² These are the proposed pathways by which we argue that the cardiovascular health of US women is increasingly deteriorating.

FIGURE 1—

Conceptual Framework of Proposed Pathways by Which the Cardiovascular Health of US Women Is Deteriorating

Note. CV = cardiovascular.

Source. Adapted from Palència et al.⁷

When CVD mortality data are further disaggregated by other inequality axes, the social gradient is magnified for all racial/ethnic groups.¹⁰ For example, during 2007 to 2011, age-adjusted CVD, heart disease, and stroke mortality rates were highest for Blacks from the most-deprived group (334.59, 131.74, and 27.47 per 100 000, respectively).¹⁰ CVD mortality rates were 1.36 (Whites), 1.32 (Blacks), 1.76 (American Indians/Alaska Natives), 1.27 (Asians/Pacific Islanders), and 1.42 (Hispanics) times higher for the most deprived group compared with the least deprived group.¹⁰ Further stratification by geographical location reveals even more pronounced associations, especially in rural populations where access to health care is limited. From 1990 to 2009, poverty gradients have steepened between rural and urban populations in the United States, after adjustment for poverty levels.¹³ The data show that premature mortality as a result of CHD contributes to the increasing rural–urban disparity and higher rural mortality.

ARE WOMEN DISPROPORTIONATELY AFFECTED?

Previous evidence suggests that women may be relatively insulated from the social gradient of health. However, a review of 36 studies from a range of countries found that the social gradient may be stronger for men than women for all health outcomes other than heart disease.¹⁴ In the United States, Singh et al. showed that the socioeconomic gradients in CVD mortality were steeper for women than men.¹⁰ In the period 2007 to 2011, women in the most deprived group had 2.5-, 2.6-, and 2.2-times higher CVD, heart disease, and stroke mortality, respectively, than did those in the most affluent group.¹⁰ Moreover, the differences in mortality observed among women from different socioeconomic backgrounds were more pronounced for CVD than for all causes.⁹ The application of an intersectionality framework reveals further disparities. Disaggregating the data further by race/ethnicity revealed that CHD mortality rates are more pronounced for Black compared with White women. For example, the age-adjusted CHD mortality rate is 99.7 per 100 000 for Black versus 80.1 per 100 000 for White women.¹⁵ Other data show that Black women living in the poorest socioeconomic conditions have the highest number of cardiovascular risk factors of all groups—a trend consistently observed across the United States. Again, the further inclusion of geographic location consistently magnified these relationships. Davis et al. found that Black women with lower socioeconomic status had worse cardiovascular health (≥ 2 self-reported risk factors) than did White women living in urban and rural parts of the southern United States.¹⁶ However, they concluded that conventional risk factors only modestly explained relative social inequality, at least in men.

HAVE INEQUALITY AXES CHANGED RECENTLY?

Over the past 50 years, women’s roles have dramatically evolved both in the United States and globally. US women now have fewer children and have them at a later age. Concurrently, there have been marked increases in education attainment and workforce participation rates. Since 1982, 1987, and 2006, a greater number of women than men have graduated with bachelor’s, master’s and doctorate degrees,¹⁷ respectively. In 2017, women represented almost 46.9% of the total workforce and 57% of women participated in the workforce.¹⁸ These shifts in social norms have afforded women greater financial independence and access to capital and resources. Paradoxically, however, 2016 US Census Bureau data show that, compared with men, women have reduced access to the total distribution of wealth and are more likely to live in poverty. Ten percent of US women aged 25 to 54 years live in poverty (based on pretax income),¹⁹ which coincides with women’s family-building stage of the life course. Indeed, poverty rates are highest for women aged 18 to 24 years (24.8%) and lowest for men aged 65 to 74 years (7.27%). Specifically, American Indian/Alaska Native (22.8%), Black (21.4%), and Latino women (18.7%) experienced the highest rates of poverty.²⁰ There is evidence to suggest that pregnancy is a trigger for income inequality. Women experience a 4% wage penalty per child, whereas men’s wages increase as a result of having children.

US women are more likely than ever before to be rearing children alone. Since 1960, the number of single-parent households has more than tripled. Women head the majority of single-parent households and more than one third of these live in poverty.²⁰ Both as a consequence and driver of living in poverty, women (particularly non-Whites) are more likely to have undocumented work status, be uninsured, live with disability, and have higher health care costs. Of major concern is the recent finding that, for those with atherosclerotic CVD, out-of-pocket annual health expenditures may be driving financial hardship in those with low income. Total out-of-pocket health care cost was higher for women ($1582) than men ($1437).²¹

Compared with men, women are subjected to restrictions on their access to wealth. Currently, the earnings ratio of US women relative to White men is 79% for White women, 62% for Black women, 54% for Hispanic women, 90% for Asian women, and 57% for American Indian/Alaska Native women, and Asian/Pacific Islander women of different backgrounds fare similarly poorly or worse than do the other groups.²² This is not simply an artifact of the (perceived) disparities in education rates. First, little progress has been made in closing the gender pay gap over the past 36 years despite female bachelor’s degree graduates outnumbering male bachelor’s degree graduates. Second, the gap narrows only slightly from 77% to 74% with increasing levels of education. Third, the gap is not explained solely by women’s full or partial absences from the workforce for the purpose of rearing children or caring for family. Importantly, the gender pay gap has consequences for cardiovascular health. As the ratio of women’s to men’s median weekly earnings becomes more pronounced across the life span, women’s risk of death from CHD increases (Figure 2).

FIGURE 2—

Women’s Coronary Heart Disease Death Rates (per 100 000) and Weekly Earnings as a Percentage of Men’s, Across Adulthood: United States, 2015

Note. CHD = coronary heart disease.

Source. Bureau of Labor Statistics, US Department of Labor, Annual Report: Highlights of Women’s Earnings in 2015: https://www.bls.gov/cps/earnings.htm; Centers for Disease Control and Prevention: death rates for diseases of heart, by sex, race, Hispanic origin, and age: United States, selected years 1950–2015: https://www.cdc.gov/nchs/data/hus/2016/022.pdf.

In examining the data by geographic regions, states with the greatest levels of inequalities appear to have the poorest CHD outcomes. In 2016, the state with both the highest cardiovascular mortality and CVD disability-adjusted life year rates,²³ Mississippi, also reported the highest rates of poverty, income inequality, and people without health insurance. Greater income inequality even by as little as 0.1 on the Gini coefficient has been shown to predict a 1% higher probability of death from CHD.²⁴ Indeed, US Census and Global Burden of Cardiovascular Diseases data show a trend toward greater CVD burden and higher poverty rates at state level (Figure 3). The ways in which the CVD burden has changed in recent decades was recently investigated by region and sex. Between 1990 and 2016, the age-standardized rate of CVD disability-adjusted life years decreased significantly in all states; a slower decline occurred for women compared with men. In some states, an increase was observed. Women residing in the Mississippi River region fared worst.

FIGURE 3—

Proportion of Population Living in Poverty and Cardiovascular Disease Disability Adjusted Life Years Rates by US State: 2016

Note. DALY = disability-adjusted life year. Data point = US state.

Source. Global Burden of Cardiovascular Diseases Collaboration (2018)²⁵ and US Census Bureau, Income and poverty in the United States: 2016: https://www.census.gov/data/tables/2017/demo/income-poverty/p60-259.html.

HOW DO WE ALLEVIATE INEQUALITY AXES?

The widening cardiovascular disparities must be considered in the context of both increasing inequalities and population life expectancy. As countries become wealthier, more progressive, and more technologically advanced, one would expect this to translate to corresponding increases in longevity for the entire population. In the main, population life expectancy improves alongside economic development in the form of gross domestic product before subsequently plateauing—a phenomenon known as the Preston Curve. In the United States, however, improvements in life expectancy have slowed more than any leading industrialized nation. Between 1990 and 2010, life expectancy declined in an unprecedented fashion—by 3.1 years for low-educated White women and 0.6 years for low-educated White men.²⁶ This is despite the United States having the world’s largest economy and per-capita expenditure on medical care. Krieger et al. calculated that had all members of the US population experienced the same yearly age-specific premature mortality rates as Whites in the highest income quintile from 1960 to 2002, 14% of White premature deaths and 30% of premature deaths among populations of color would not have occurred.²⁷

Alongside the declining life expectancy rates is the concurrent increase in maternal mortality rates—a key indicator of both population health and gender equality. Maternal mortality in the United States is the highest among industrialized nations (26.4 per 100 000 live births)²⁸ and shows significant racial/ethnic disparity. Black women are at highest risk for maternal death compared with White women and women from other racial/ethnic backgrounds (40 vs 12.4 vs 17.8 per 100 000, respectively). Importantly, cardiovascular-related maternal deaths (CVD, cardiomyopathy, cerebrovascular accident, hypertensive disorder) contribute the biggest burden, accounting for 39.7% of all pregnancy-related deaths.²⁸

In the absence of corrective measures to alleviate widening inequality axes, reductions in life expectancy and deteriorations in cardiovascular outcomes will likely continue, women with lower levels of education being most vulnerable. To date, the most commonly targeted areas of health equity policy have been social protection and welfare, health systems management and health insurance, environment and living conditions, and taxation. Economists have argued that the formulation of government policies that promote equitable distribution of the national income is essential. This would require that all forms of labor be counted. Paradoxically, the unpaid labor of women—who spend more time doing unpaid work than men and are the key beneficiaries of investment in welfare and social services that comes from economic prosperity—is not recognized as part of gross domestic product.

Importantly, women experience higher morbidity and mortality in states in which they have lower political participation and economic autonomy. Affirmative action on gender, race, and other inequality axes to promote parity in all aspects of society including workplace representation, conditions, and security; education; and health service access is required to correct these unprecedented trends in CVD mortality in premenopausal women. Indeed, greater material, political, and psychosocial empowerment is a potent predictor of improvements in one’s living conditions and, therefore, health. Perhaps the most widely discussed policy has been the provision of universal health care. While critical to the health and vitality of a population, affordable health care is 1 aspect of the redistribution of wealth. Since 1982, Alaska has been providing universal basic income via the Alaska Permanent Fund Dividend.²⁹ Data show that while child poverty rates in Alaska have been increasing, poverty rates have substantially declined by 46% from 2000 to 2015 for rural Alaska Indigenous persons.

More recently, the introduction of the Affordable Care Act (ACA) has had profound effects on access to affordable health care for US citizens. The data show that women aged 19 to 64 years have in fact made the greatest insurance coverage gains of any group.³⁰ Before the introduction of the ACA in 2010, data show that one third of women who attempted to purchase health plans on their own were refused, charged an excess premium related to their health status, or had health problems excluded from plans. Insurance reforms in 2014 facilitated coverage for millions of women without employer insurance. Findings from the Commonwealth Fund Biennial Health Insurance Survey, in 2016,³⁰ showed not only that women aged 19 to 64 years obtained coverage more easily, with the proportion of those deferring required care because of costs plummeting to an all-time low (from 20% uninsured in 2010 to 11% in 2016), but also that insurance status predicted engagement with preventive health care such as mammograms.

It stands to reason that this would extend to cardiovascular care, cardiovascular screening, and other programs that directly and indirectly shape women’s short- and long-term CVD incidence and recurrence. Indeed, these data may help explain, for example, the increase in incident myocardial infarction admission for young Black women since the implementation of ACA. Greater coverage of previously uninsured or underinsured patients would likely encourage hospital presentation. However, analyses of National Hospital Ambulatory Medical Care Survey data from 2006 to 2013 found, that while ACA has promoted cardiovascular preventive care recommended by the US Preventive Services Task Force, including diabetes, hypertension, and tobacco screening, sex disparities remain with respect to aspirin therapy intitiation.³¹ This infers that removal of structural barriers alone is insufficient for women to receive evidence-based cardiovascular preventive care if biases exist at the provider level.

Given that women are disproportionately affected by the social gradient of CVD and inequality axes, corrective action that goes beyond health insurance coverage could improve their health, well-being, and exposures that elevate their cardiovascular risk. For example, a US population-level study investigated the relationship between the gender pay gap and violence against women, measured through female hospitalizations for assault.³² The authors found that a reduction in the gender wage gap explained 9% of the decline in domestic violence from 1990 to 2003.³² Yet, it is currently unclear whether mainstreaming gender and its intersecting inequality axes can offset the excess CVD burden observed in young women. That is, comprehensive analyses of community-, household-, or family-level interventions that could have an impact upon women’s cardiovascular health would need to quantify the extent to which gender-equitable practices themselves (as distinct from the other residual benefits such as increased access to quality health care) were responsible for improving cardiovascular outcomes. Many of the examples from other countries such as Finland or Sweden are conducted in a setting in which some of the smallest social inequalities are observed to begin with because of their egalitarian foundations. It is promising that this evidence shows that targeting key indicators of gender equality can improve both women’s and men’s health. By investigating trends in gender inequality and cardiovascular deaths at the state level, preliminary data show that greater gender equality may be associated with lower CVD mortality rates (Figure 4).

FIGURE 4—

Gender Equality and Cardiovascular Disease Mortality (Female), by US State: 2017

Note. CVD = cardiovascular disease.

Source. Centers for Disease Control and Prevention (CDC) National Vital Statistics System, 2013–2015. Data used to create Gender Equality Ranking (2018) were generated using 16 indicators collected from the US Census Bureau, Bureau of Labor Statistics, Equal Employment Opportunity Commission, National Women’s Law Center, National Center for Educational Statistics, CDC, and Center for American Women and Politics.

The health benefits associated with investing in gender equality initiatives and affirmative action are not confined to the micro level. Gender equality at the macro level results in improvements in everyone’s health behaviors (e.g., alcohol consumption). The economic projections of the health benefits of public spending at local and state level are impressive. Each additional $250 per capita spent on welfare annually predicted a lower probability of dying from any cause (3%) and CHD (1.6%). In 2019, a National Heart, Lung, and Blood Institute Workshop Report detailed its recommendations for intervention strategies to alleviate widening disparities in CVD health.³³ Within this context, understanding the ways in which key biological mechanisms that are driven by and perpetuate inequality axes in cardiovascular health will be critical.

WHAT CAN BE DONE WHILE WE AWAIT SOCIAL REFORM?

The roles of health care, health services, and health professionals are critical in the context of reducing cardiovascular disparities within a social determinants and intersectionality framework. System- and individual-level biases—shaped by and that perpetuate inequality axes—affect one’s receipt of care at multiple levels. Biases underpin policymaking by insurers and government regarding who is deemed eligible to access health care and can also influence the quality of individual patient care. Since the early 2000s, myocardial infarction hospitalization rates have increased for younger Black women compared with White women.³⁴ There is also evidence that young, Black, rural women especially are more likely to receive suboptimal health care after myocardial infarction admission.³⁴ While clinicians can take a bottom-up approach—for example, providing risk assessment and health care that considers intersectionality—currently, such tools for use in clinical settings are few and far between.³⁵

Transforming clinical care for women ultimately requires deviation from the largely male biomedical model of care. An example of this is the Women’s Heart Centers, which provide specialized, comprehensive care for women across the United States.³⁶ Lundberg et al.³⁷ detail the “ideal elements” of clinical care across the life span for US women from a diverse range of cultural-linguistic, socioeconomic, and geographical backgrounds. The constituents that most pertain to CVD health include

• multidisciplinary and integrated medical team approach (e.g., CVD and pregnancy, cognitive and mental health, nutrition, and exercise);

• specialized patient intake forms with detailed historical and screening questions pertinent to CVD in women (e.g., reproductive history, migraines, rheumatologic disorders) that include CVD risk factor information;

• addressing special needs of women with genetic disorders (e.g., adult congenital heart disease, familial hypercholesterolemia);

• care models that address gender roles of, and social determinants of, CVD for women; and

• expertise in cardiovascular diseases unique to, or seen more often in, women (e.g., spontaneous coronary artery dissection).

Similarly, for cardiovascular prevention, tailored risk reduction programs for women of low income such as WISEWOMAN (Well-Integrated Screening and Evaluation for Women Across the Nation)—which uses community- and health system–based approaches—have been formally evaluated with impressive results across a multitude of settings.³⁸

THE ROLE OF WOMEN AS CITIZENS

Private insurers, health care providers, and public policy makers—and their structures—are undoubtedly accountable to citizens. Citizen science and participatory research are critical to the development and delivery of effective interventions and policies that are codesigned by communities often underserved by conventional public health systems.³⁹ These approaches are not without limitations. They often employ a binary approach to sex and gender, fail to address the nuances of inequality axes, and are implemented by those with limited capacity to act as change agents. Going forward, this is an important consideration.

FUTURE RESEARCH

To this end, gender- and intersectional-sensitive tools that have capacity to appropriately and accurately monitor and evaluate such interventions are required. Such tools require process, impact, and outcome evaluation metrics to enable identification and better understanding of the mechanisms by which corrective initiatives can have an impact upon women’s health, where currently these are lacking. More broadly, there are well-documented limitations of using population-based surveillance data to disentangle drivers of disease in subpopulations and the application of intersectional frameworks in clinical and research settings. There are also few rigorously designed randomized controlled trials that unequivocally show that interventions designed to alleviate the axes of inequality improve cardiovascular outcomes. There are several reasons for this; it is likely that interventions that occur at the structural level take decades to affect the cardiovascular health of individuals, and quantifying cause and effect is difficult. Moreover, interventions that attempt to target 1 axis rather than multiple inequality axes are more likely to fail or at least dilute their true effects.

In conclusion, the impressive reductions observed in CVD mortality over recent decades have not been afforded to the entire US population. Population-level data obscure stark inequalities in CVD health, particularly as they affect premenopausal women who are subject to an increasing number of inequality axes. Going forward, CVD prevention strategies need to address inequality axes both in the context of CVD population health approaches and the provision of tailored care. These should be developed with communities using citizen science and participatory frameworks as a national priority for CVD prevention and control. Moreover, formal evaluations into the impact of corrective action such as gender equity initiatives on the cardiovascular health of subpopulations of women using an intersectional framework is required.