Racial Disparities in Cardiovascular Health Among the Acute Coronary Syndrome Population

Abstract

The relative distribution of proportions of cardiovascular health (CVH) categories within racial groups has been examined. However, little scientific evidence exists on the gap trend in racial/ethnic disparities in mean CVH score among non-Hispanic (NH) Whites and Blacks. This study examined the trend(s) in the gap(s) in predicted CVH scores between NH Whites and Blacks over 10 years. In a cross-sectional analytical study, 10 years of Medical Expenditure Panel Survey data from 2008 to 2018 were pooled, utilizing multivariate Poisson's regression of CVH metrics on race, while controlling for relevant covariates. The interactions of acute coronary syndrome (ACS) with CVH metrics, and other key variables such as trends and grouped Charlson Comorbidity Index allowed for variations in the effect of these variables on the subgroups. The mean gap in CVH scores was on average 0.15 [95% confidence interval (CI) 0.137 to 0.170], with Blacks consistently having reduced odds of having ideal CVH until 2014. The overall impact of having an ACS decreased acquired CVH scores by 24.1% [95% CI −0.275 to 0.207], and was equal for both racial subgroups (P < 0.05). The Affordable Care Act (ACA)-trend was positive, increasing the likelihood of improved CVH in the sample (P < 0.05), deflecting a downward trend in acquired CVH scores for both races, as the gap narrowed into more recent years. The CVH gap was stabilized by the ACA, but never really converged, suggesting that efforts to reduce existing disparities between Blacks and NH Whites in the United States would require government policies to look beyond mere “access” and/or “affordability” to health care.

Introduction

Racial and ethnic disparities in cardiovascular morbidity and mortality are often to the disadvantage of minority groups.^1,2 The effectiveness of policy interventions aimed at decreasing cardiovascular health (CVH) disparities relies on studying the gap trend over the years and exploring the factors underlying the racial health gap between minorities and prevalent White populations.

Despite a drop in cardiovascular mortality over the past 50 years, cardiovascular disease (CVD) has remained the leading cause of death in the United States across all racial and ethnic groups.³ The burden of CVD exhibits notable racial and gender disparities, with non-Hispanic (NH) Black individuals experiencing higher rates of coronary heart disease, heart failure, stroke, and overall mortality from CVD when compared with NH White individuals.^3,4

CVD and its related risk factors are not evenly distributed among different racial and ethnic groups, indicating a disproportionate burden. There are wide geographic and racial variations in mortality, particularly for ischemic heart disease and stroke. Men have higher rates of CVD compared with women within the same race/ethnicity groups.

Public health initiatives such as the Healthy People 2020 and the current American Heart Association (AHA) Strategic Impact Goals aim at eliminating health disparities in this area.^5,6

Disparities in CVH based on race and ethnicity could potentially translate into disparities in the incidence of CVD in the future. Previous studies have repeatedly demonstrated that NH Whites have a higher average number of ideal health factors and behaviors compared with aggregated NH Blacks and Hispanics.⁷ Understanding of how racial and ethnic patterns in this context may have evolved over the past decade is limited.

This paper will assess the racial/ethnic disparities in average CVH scores between NH Whites and Blacks between 2008 and 2018, emphasizing whether “greater access” to preventative care influenced existing differences in achieving ideal health scores. Given the nation's greater access to preventive care made possible by the Patient Protection and Affordable Care Act (ACA), it is expected that the racial gap in average CVH scores would narrow following its implementation during the period studied.

According to the AHA, achieving good CVH is related to at least 3 health behaviors and 3 health factors called the CVH metrics, which form the basis for the AHA's “Life's Simple 7” (LS7).⁸ The behavioral factors include inadequate physical activity, obesity, smoking, and health factors are high cholesterol, hypertension, and diabetes mellitus.

Racial disparities in cardiovascular health

Some articles show how the concept of CVH is utilized to assess the likelihood of event occurrences and existent disparities in chronic disease burden⁹ focused on comparing both health outcomes and CVH scores between White and Black Americans. Brown et al¹⁰ found that Blacks demonstrated higher stroke mortality rates than Whites because of a greater prevalence of comorbid conditions such as hypertension, diabetes mellitus, and obesity. These findings, specifically among those without prevalent CVD over a 26-year-time period, agree with what other researchers discovered.¹⁰

While the proportion of ideal CVH management scores among Whites remained slightly below 40%, rates of optimal scores among Mexican and Black Americans were considerably lower at 25% and 15%, respectively. The gap in optimal CVH was generally highest for Black Americans relative to Whites.¹⁰ Ogunmoroti et al¹¹ compared the racial disparities in heart failure with the CVH metric scores and found out that low CVH scores mediated higher incidence of heart failure and non-cardiovascular (CV) disease compared with ideal CVH scores.

Yet, though the gap between Blacks and Whites of those with ideal CV-health conditions persisted over time, a similar somewhat narrower gap existed between Mexican Americans and Whites as it narrowed down to 8% among younger age groups (<44 years). But this converging was due to 15% points drop in ideal CVH among Whites. Thus, there had not been substantial improvement in CVH among African Americans and Mexican Americans, but instead a general drop in CVH among Whites.¹⁰

A similar trend was discovered in gender comparisons by Pool et al¹² when they explored disparities in the overall CVH for NH Black and Mexican American women versus NH White women between 2000 and 2012. NH Black females or Mexican American females showed considerably lower average CVH scores compared with NH White females in virtually all survey cycles, but NH White females did regress in their CVH in more recent years.

Methods

Study design

This study was an analytical cross-sectional study using secondary data from 2008 to 2018 from the Medical Expenditure Panel Survey (MEPS) database. The MEPS is a set of large-scale national surveys of households, individuals, Insurance companies, health providers (doctors, pharmacies, and hospitals), and employers across the United States, sponsored by the Agency for Healthcare Research and Quality.

The panel consists of randomly sampled noninstitutionalized US civilians providing nationally representative estimates of sociodemographic characteristics, medical conditions, and utilization and costs.¹³

Study population

The team included adults aged 40 and older with and without acute coronary syndrome (ACS) who participated in the MEPS between 2008 and 2018. To obtain the final sample, 2 MEPS components were merged: the household component, which includes data on demographic characteristics, and the medical conditions component, which includes a “current” condition per person at any time during the data year.

Pooling the merged data for 10 years from 2008 to 2018 allowed for a larger and analyzable population of adults with and without ACS. The year 2017 was excluded because the variable for body mass index (BMI), an essential component of CVH metrics, could not be created for that year.

Individuals with ACS were identified for the years 2008–2015 using the International Statistical Classification of Diseases, Clinical Modification (ICD-9 CM) diagnosis of the condition, and ICD-10 CM for the years 2016 and 2018. Charlson's Comorbidity Index was generated using the same ICD-9 and ICD-10 CM codes used to specify medical conditions. ACS refers to the self-reported history of diagnosis of unstable angina and acute myocardial infarction, ST-elevation (elevation of ST segment on electrocardiogram) myocardial infarction, or non-ST elevation myocardial infarction.

Main independent variable

The primary independent variable of interest was race. Self-reported race was measured to operationalize race or ethnicity. There were 2 race categories: NH Whites and Blacks.

Primary outcome

The primary outcome variable was the mean difference in CVH metrics between NH Whites and Blacks. The team generated the CVH metric variable as a linear term by adding up all individual CVH components—physical activity, smoking status, BMI, blood pressure, blood glucose, and total cholesterol—together into an aggregate while granting them equal weights.

Non-dietary cardiovascular health

The primary outcome variable of interest was the average CVH score count for each participant, represented by the CVH management score, with a range of 0 to 6. The non-dietary CVH components examined in this study included inadequate physical activity, obesity, smoking, hypercholesterolemia, hypertension, and diabetes mellitus. Diet was not assessed in MEPS, and therefore, the maximal count of 6 was achieved as a composite score, as did another research.¹⁴

Participant responses from the self-administered questionnaire were utilized to determine the average CVH score of participants, classifying everyone with a binary variable (favorable [0] vs. unfavorable [1]) for the CVH metrics. Those who reported a diagnosis of a cholesterol disorder, hypertension, diabetes mellitus, smoked within <1 year of the time of the interview, did not engage in moderate vigorous physical activity 5 times a week, or had a BMI >25 or <18 kg/m² were classified as having an unfavorable risk factor.

They were then assigned a 0, whereas those who did not exhibit any of these adverse pre-conditions were assigned a value of 1. The sum of values assigned to determine a participant's CVH score (or LS7 components) ran from 0 to 6, with six being the ideal state of CVH and 0 being the poorest state of CVH. Based on the achieved level of CVH scores, survey participants were categorized as having optimal (≥5 CVH scores), average (2–4 CVH scores), or poor (0–1 CVH scores).

Individual comorbidity level was assessed using the grouped Charlson Comorbidity Index (GCCI), which is known to be a method of categorizing comorbidities of patients based on the ICD diagnosis codes found in administrative data, such as hospital abstracts data.¹⁵ It stratifies comorbid conditions by disease severity using associated weights (from 0 to 6), based on the adjusted risk of mortality or resource use (Concept: Charlson Comorbidity Index, n.d.).¹⁶

Covariates

The authors also examined age, education level, insurance status, sex, race/ethnicity, and family income as factors that tend to be associated with CVH scores, and therefore were explored as covariates in the determination of the disparities in CVH in both ACS and non-ACS cohorts that were White or Black. Race/ethnicity (Black, and White), sex (self-identified as male, self-identified as female, etc), age (in years), and region (NEast, MWest West, South) were self-reported and collected for each MEPS participant.

Family income level (very low income: <125% of the federal poverty level [FPL]; low income: 125% to <200% of the FPL; middle income: 200% to <400% of the FPL; or high income: >400% of the FPL), educational level (having less than a bachelor's degree, having achieved an equivalent of a bachelor's or master's degree and then a Doctoral or professional degree).

Statistical analysis

The team pooled 10 years of MEPS data from 2008 to 2018 using descriptive statistics to summarize frequency distributions with corresponding weighted proportions by ACS versus non-ACS status, after which they merged the medical conditions with the demographics file using key variables. First, they ran a Poisson's regression of CVH on the interaction of Race with ACS in an unadjusted regression to allow for variation in the racial gap in CVH scores based on ACS status, while accounting for age, sex, race/ethnicity, region, insurance status, level of income, education levels, and Charlson Comorbidity Index.

Charlson's Index was used to explore the impact of comorbidity burden on the associations of CVH score status with racial background. In a second step, the team utilized trend variables for ACA to determine the impact of ACA on existential gap trends.

Cells contained coefficients and 2-sided P values indicating statistical significance with P < 0.05. To estimate the average CVH scores for Whites and Blacks with and without ACS, the regression model would be Poisson's regression.

Ln (CVH-score)_i = β₀ + β₁ (ACS × White)_i + β₂ (ACA × trend)_i + β₃ (GCCI)_i + β₄ (male)_i + β₅ (I.educlevel)_i + β₆ (trend)_i + β₇ (I.age)_i + β₈ (I.insurance)_i + β₉ (West)_i + β₁₀ (MWest)_I + β₁₁ (NEast)_i +β₁₂ (I.incomcat)_i + e_i.

Where

• CVHscore_{meani i}: represents the average difference in CVH scores between both races.

• I.age _i: represents the patient's age.

• I.incomcat _i: represents income category based on FPL.

• GCCI: Grouped Charlson's comorbidity Index.

• Male_i: 1 if subject self identifies as male, 0 otherwise.

• I.educlevel_i: indicates level of highest education attained.

• ACA_trend_i: Interaction of the ACA with time trend variable.

• White_trend_i: Interaction of White with trend variable.

• West_i: represents the variable for region (West, NEast, Midwest).

• ACS_I: ACS indicator.

• e_i: Error term with normal distribution.

Results

Characteristics of the sample

The sample consisted of 749,180 MEPS participants across 2008–2018, who were more than 40 years of age, with a BMI ranging between 18 and 25 kg/m², for normal BMI, and who either had or did not have ACS. The mean age for the ACS subgroup was 67.5 years (standard deviation [SD] 9.2) and 47.2 years (SD 2.2) for the non-ACS subgroups, with 59.8% being female. The distribution of means of proportions in demographic characteristics for the study population between Whites and Blacks is shown in Table 1.

Table 1.

Descriptive Statistics of Cohorts

	ACS, % (95% CI)	Non-ACS, % (95% CI)	Difference (P values)	Combined sample, % (95% CI)
Prevalence
Whites	80 (79.4–80.6)	79.2 (79.21–79.3)	<0.001	79.2 (79.1–79.3)
Blacks	20. (19.4–20.5)	20.7 (20.7–20.8)	<0.001	20.7 (19.4–20.5)
Male	56.7 (56.0–57.4)	39.9 (39.8–40.0)	<0.001	40.2 (40.1–40.3)
US region
West	16.7 (16.2–17.2)	22.5 (22.5–22.7)	<0.001	22.5 (22.4–22.6)
MWest	21.9 (21.3–22.5)	22.9 (22.8–23.0)	<0.001	23.0 (22.8–22.9)
NEast	17.3 (16.7–17.8)	16.0 (15.9–16.1)	<0.001	16.0 (15.9–16.1)
Insurance status
Private	46.9 (46.2–47.6)	55.3 (55.2–55.4)	<0.001	55.2 (55.1–55.3)
Medicare/Medicaid	48.6 (47.9–49.3)	36.8 (36.7–36.9)	<0.001	36.9 (36.9–37.0)
Age groups, years
40–64	51.0 (50.2–51.9)	43.4 (43.3–43.5)	<0.001	43.5 (43.4–43.6)
65–79	44.3 (43.6–45.0)	19.2 (19.1–19.3)	<0.001	19.6 (19.5–19.7)
>80	21.0 (20.5–21.6)	7.4 (7.3–7.5)	<0.001	7.7 (7.6–7.7)
Family income category (FPL)
<125% FPL	27.3 (26.7–27.9)	25.6 (25.4–25.6)	<0.001	25.6 (14.6–14.7)
125%–200% FPL	16.2 (15.7–16.7)	14.6 (14.5–14.7)	<0.001	14.7 (14.6–14.7)
200%–400% FPL	25.8 (25.2–26.4)	25.7 (25.5–25.9)	0.6574	25.7 (25.5–25.7)
>400% FPL	30.6 (30.0–31.2)	34.1 (34.0–34.2)	<0.001	25.7 (34.0–34.2)
Highest educational level attained
Bachelor's/Master's degree	17.1 (16.6–17.7)	20.2 (20.1–20.3)	<0.001	20.1 (20.0–20.3)
PhD or Professional Degree	2.7 (2.5–2.9)	3.1 (3.0–3.1)	<0.001	3.1 (3.06–3.16)
Sample size	20,885	1,225,191	—	1,246,076

Means and proportions were compared across all groups.

ACS, acute coronary syndrome; CI, confidence interval; FPL, federal poverty level.

Of the ∼749,180 adults in 2008–2018, 95.82% did or did not have an ACS event; 80% self-identified as being NH White and having had at least 1 event of ACS. Meanwhile almost 20.7% of individuals without a history of ACS reported being Black, and about 20% of the entire ACS population were of Black race. While 40% of the sample were male, about 56% of those affected by ACS were male.

The proportions of patients with ACS events looked somewhat evenly distributed across all regions in the United States, with the Mid-West having the most representation (21.9%). The proportion of subjects who were privately insured and had an ACS event was slightly lower than those with public insurance, ranging from 46.6% to 48.6%, respectively. Meanwhile the incidence of ACS was almost inversely associated with age at time of the interviews; median household income seemed to be directly proportional to the advent of ACS.

In the first set of unadjusted regressions (Table 2), the team observed that, on average, across all 11 years, NH Whites had an almost 15.3% [95% confidence interval (CI) 13.7% to 17.0%] higher mean CVH scores compared with Blacks. That is to say, if a Black individual had a CVH score of 3.0, his White counterpart with similar characteristics would have an average score of 3.45.

Table 2.

Results for the Prediction of the Mean Number of Cardiovascular Health Scores Based on Race and Acute Coronary Syndrome Status for the Years 2008–2018 Using the Medical Expenditure Panel Survey Dataset

Mean CVH (0–6)	Unadjusted specifications (95% CI)	Adjusted specifications (95% CI)	Adjusted specifications + trends (95% CI)
Constant (NH Black is Reference)	0.969a (0.954 to 0.985)	0.981a (0.955 to 1.01)	1.017a (0.987 to 1.04)
NH White	0.153a (0.137 to 0.170)	0.100a (0.083 to 0.113)	0.094a (0.069 to 0.120)
Trend (2008 = 0)	—	—	−0.015a (−0.022 to −0.009)
White # Trend	—	—	0.002 (−0.006 to 0.010)
ACA # Trend	—	—	0.036a (0.028 to 0.044)
(ACA or year ≥2014)	—	—	−0.205a (−0.250 to −0.160)
ACA # Trend # White	—	—	−0.002 (−0.009 to 0.004)
ACS	−0.241a (−0.275 to 0.207)	−0.183a (−0.202 to −0.165)	−0.190a (−0.217 to −0.164)
White # ACS	−0.000 (−0.040 to 0.038)	—	—
GCCI	—	−0.029a (−0.040 to 0.0192)	−0.029a (−0.040 to −0.019)
Male	—	−0.088a (−0.100 to 0.077)	−0.088a (−0.099 to −0.077)
Bachelors or Masters	—	0.091a (0.078 to 0.104)	0.101a (0.088 to 0.114)
Terminal degree/PhD	—	0.147a (0.122 to 0.174)	0.172a (0.148 to 0.197)
Northeast	—	0.050a (0.032 to 0.069)	0.050a (0.032 to 0.069)
Mid-West	—	0.027a (0.007 to 0.048)	0.028a (0.008 to 0.050)
West	—	0.089a (0.071 to 0.109)	0.089a (0.070 to 0.108)
Age 65–79	—	−0.105a (−0.120 to 0.092)	−0.103a (−0.117 to −0.089)
Age >80	—	0.018 (−0.003 to 0.040)	0.011 (−0.024 to 0.019)
Family income: 125%–200% FPL	—	0.033a (0.014 to 0.054)	0.032a (0.012 to 0.052)
Family income: 200%–400% FPL	—	0.090a (0.074 to 0.107)	0.088 (0.071 to 0.104)
Family income: 400% + FPL	—	0.159a (0.142 to 0.177)	0.154a (0.137 to 0.172)
Medicare/Medicaid	—	−0.133a (−0.159 to 0.108)	−0.133a (−0.159 to −0.108)
Private insurance	—	−0.037 a (−0.062 to 0.013)	−0.038a (−0.062 to −0.013)

Standard error in parenthesis.

^a Signifies 5% significance 2-sided hypothesis test.

ACA, Affordable Care Act; ACS, acute coronary syndrome; CI, confidence interval; CVH, cardiovascular health; FPL, federal poverty level; GCCI, Grouped Charlson comorbidity Index; NH, non-Hispanic.

The mean CVH score for NH Whites was 3.0 and 2.64 for Blacks. The overall impact of having acquired ACS decreased acquired CVH scores by 24.1% [95% CI −27.5% to 20.7%]. So, the mean difference between NH Whites and Blacks who have acquired ACS is still 15.3%, given that the impact of ACS is the same for everybody. This is because the differences in the impact of ACS between both races were not significant at the 5% level.

The other controls had some interesting findings as well. Three relative income controls were based on reported income divided by the FPL. Those with more income consistently displayed higher CVH scores than those who were at the FPL. Meanwhile, those with more than 2.5 times the FPL had a 9% higher CVH score on average, relative to those at the poverty level in income; individuals having an income level well above poverty tended to experience the highest CVH sores, that is, they had a 16% [95% CI 14.2% to 17.7%] higher CVH score than those at the FPL.

The GCCI controlled for the impact of serious illnesses (or the co-morbidity burden) and tended to decrease CVH scores by 3.0% [95% CI −4.0% to 1.92%] per unit change in Charlson's Index. Males had, on average, about 9% [95% CI −10.0% to 7.7%] lower CVH score than females. The population associated with public insurance (Medicaid and Medicare) showed a more significant decrease in mean CVH score on average, percentagewise, compared with privately insured participants regarding how far their scores were above the mean scores of the uninsured population.

A college education, or greater, predictably statistically significantly increased the predicted level of CVH scores and tended to increase the probability of having an ideal CVH. Meanwhile, participants from all regions other than the south demonstrated better ability to improve their CVH scores, that is, engage to a lesser extent in risky CV behavior; those in the western region showed the most ability to achieve ideal CVH scores (P < 0.5). The P values of these estimates were significant at the 5% level of significance.

The impact of the trend variable was to increase the frequency of participants engaging in behavior that will ultimately put them at greater risk of developing CVD. Meanwhile, the trend showed a yearly decline of sustenance of health behavior by 1.5% [95% CI −2.2% to −1.0%]; the White race showed a yearly 0.2% of higher abilities to improve overall CVH. That would imply that extrapolated over 10 years, there would be a persisting 2% difference (gap) in CVH score between NH Whites and Blacks.

The role of the ACA

Because of the ACA, comprehensive coverage of tobacco cessation offered a new opportunity to encourage tobacco cessation, to minimize tobacco use.^5,17 This was because the full implementation of the ACA in 2014 assisted current smokers to quit smoking, and reduce obesity, among other modifiable risk factors. It was, therefore, interesting to see the impact of the interaction of ACA with the trend variable on CVH metrics and race on the sample.

In this study, the authors saw that even though the ACA-Trend was positive, that is, it increased the likelihood of acquiring better CVH in the sample (P < 0.5), the interaction effect of an ACA-Trend variable with White race was not statistically significant. This seemed to suggest that the impact of the ACA-Trend was not to reveal racial differences in CVH scores. In other words, there were no significant differences on how the ACA impacted both racial trends in acquired CVH scores. Figure 1 shows how before the implementation of the ACA in 2014, there was a downward trend in acquired average CVH scores for both races, but more so for the White race as the gap narrowed into more recent years.

FIG. 1.

Results of disparities in Cardiovascular Health Scores between non-Hispanic Whites and Blacks in the United States for the years 2008–2018 using the Medical Expenditure Panel Survey dataset.

These findings were similar to the results of other scientists, who found out that, though the gap between Blacks and Whites of those with ideal CVH conditions persisted over time and a similar somewhat narrower gap existed between Mexican Americans and Whites, it has narrowed down to 8% among younger age groups (<44 years).

But this is due to 15% points drop in ideal CVH among Whites. Thus, there had not been substantial improvement in CVH among African Americans and Mexican Americans, but instead a general drop in CVH among Whites.¹⁰ A similar trend was discovered in gender comparisons by Pool et al¹² when they explored disparities in the overall CVH for NH Black and Mexican American women versus NH White women between 2000 and 2012.

NH Black females or Mexican American females showed considerably lower average CVH scores compared with NH White females in virtually all survey cycles, but NH White females did regress in their CVH in more recent years.

According to Figure 2, an extrapolation of the lines for both NH Whites and Blacks would indicate a convergence down the line, meaning the differences in mean CVH would even out down the line. However, the impact of the ACA trend variable was to improve CVH scores for both races and perhaps to avert the gap getting worse over time.

FIG. 2.

Results of the impact of ACA on the trends in mean differences in ideal cardiovascular health scores between Whites and Blacks in the United States for the years 2008–2018 using the Medical Expenditure Panel Survey dataset. ACA, Affordable Care Act.

Figure 3 shows that initially, the probability of getting into the poor category was rising for both groups, but slightly faster for NH Whites. The starting off gap was about 0.06 in 2008 and rose to 0.09 in 2013. However, after the ACA was implemented, the team noticed a deflection in the trend, returning to 0.06 and staying constant. This suggests a sense in which the ACA might have deflected the rising trend, and hence widening gap in poor CVH scores, and restoring an even gap.

FIG. 3.

Results of the impact of ACA on the trends in mean differences in poor cardiovascular health scores between non-Hispanic Whites and Blacks in the United States for the years 2008–2018 using the Medical Expenditure Panel Survey dataset. ACA, Affordable Care Act.

From Table 3, the reader can see that the proportion of individuals dropping down to the poor category is 6.2% [95% CI −8.4% to −4.0%] lower for Whites than for Blacks. On the other hand, the NH Whites still have the edge over Blacks when it comes to the rate at which individuals join the ideal categories, namely 9.2% [95% CI 7.8% to 10.7%] (P < 0.5).

Table 3.

Results for the Prediction of Being in the Ideal Versus Poor Category of Achieved Cardiovascular Health Scores Based on Race for the Years 2008–2018 Using the Medical Expenditure Panel Survey Dataset

Predicted probability	Ideal CVH (5–6)		Poor CVH (0–1)
Specifications	Basic (95% CI)	Basic + Trends (95% CI)	Basic (95% CI)	Basic + Trends (95% CI)
Constant (Black is Reference Group)	0.076a (0.067 to 0.085)	0.080a (0.069 to 0.091)	0.197a (0.179 to 0.216)	0.183a (0.163 to 0.204)
White, no other race or ethnicity	0.090a (0.076 to 0.101)	0.092a (0.078 to 0.107)	−0.068a (−0.088 to 0.048)	−0.062a (−0.084 to −0.040)
Trend (2008 = 0)	−0.001 (−0.001 to 0.001)	−0.001 (−0.005 to −0.002)	0.002 (−0.000 to 0.005)	0.011a (0.005 to 0.017)
White # Trend	−0.001 (−0.003 to 0.008)	−0.004 (−0.008 to 0.000)	−0.000 (−0.004 to 0.002)	−0.004 (−0.012 to 0.003)
ACA # Trend	—	0.0078a (0.003 to 0.012)	—	−0.018a (−0.025 to −0.011)
(ACA or Year ≥2014)	—	−0.051a (−0.082 to −0.021)	—	0.087a (0.049 to 0.126)
ACA # White # Trend	—	0.002a (−0.001 to 0.006)	—	0.003a (−0.002 to 0.010)
R-squared	0.0052	0.0057	0.0040	0.0051

^a Signifies 2-sided 5% significance levels.

ACA, Affordable Care Act; CI, confidence interval; CVH, cardiovascular health.

Meanwhile, the initial trend was to raise the acquisition of CVH by 1.1% [95% CI 0.5% to 1.7%] per year, the impact of the ACA was to reverse that trend by causing the rate at which participants left the poor category to be 1.8% [95% CI −2.5% to −1.1%] per year. However, the trend for Whites leaving the poor category was about 0.3% slower than Black participants.

Discussion

To the authors' knowledge, this is 1 of the first studies to examine racial disparities among NH Whites and Blacks between 2008 and 2018 using a nationally representative sample. The authors observed that despite the increased attention given to reducing racial and ethnic disparities in public health, there has been no significant improvement in narrowing of overall CVH scores between 2008 and 2018. This means that disparities in CVH between the 2 racial and ethnic groups have persisted over the past decade, with little indication of any reduction in these differences despite the implementation of the ACA.

Utilizing the MEPS and both multivariate Poisson regression and linear probability regression analysis, the team observed time-varying disparities in CVH score between NH Whites and Blacks. This finding contrasts with how no interaction effects were found between race and ACS status, implying that changes in CVH habits following an ACS event did not differ by race, apart from differences accounted for by other covariates.

The results showed that socio-economic indicators, such as the highest level of education attained, were positively associated with a higher CVH management score. Similarly, individuals living in the south demonstrated the lowest CVH score categories, controlling for other characteristics, consistent with previous findings of Gebreab et al, whose study showed that there were higher rates of poor CVH and CVD mortality clustered in the southern states.¹⁸

In a second set of analysis, trend variables were used to explore variations in differences in CVH scores between both races. The authors observed significant variations in CVH across years. First, in an unadjusted multiple linear regression, they showed that on average, NH Whites had slightly better mean CVH scores than Blacks, and after accounting for multiple covariates, the differences were still present, though somewhat smaller. Acquiring ACS made it such that there was no significant racial disparity in CVH scores. The drop in CVH was equal in both racial groups.

While the trend for mean CVH scores was on a downward trajectory across all 10 years, the advent of the ACA offset that trend by causing improvements in average scores by 3.6% per year across all 10 years, consistent with the findings of Tabb.¹⁹ In all aspects of the analysis, things were tending to get worse before the adoption of the ACA and then started to get better when it was implemented in 2014, although the racial gap remained stable, though not getting worse.

Several prior studies have also identified variations in CVH by race/ethnicity, with NH Whites demonstrating higher average numbers of ideal health factors and behaviors compared with NH Blacks and Hispanics in aggregate.^7,20,21 One specific article that examined trends in the United States over time with regard to the mean number of ideal factors and behaviors within different racial/ethnic categories showed persistent disparities between NH Whites and both NH Blacks and Hispanics.²²

This study contributes to the understanding of long-term CVH trends by providing a breakdown of racial/ethnic differences by gender, and by assessing CVH across the full spectrum using the 6-point CVH composite score, which evaluates ideal, intermediate, and poor CVH. The authors discovered that the ACA brought improvements in overall CVH scores to both NH Whites and Blacks to the same degree; however, it failed to create convergence.

A major strength in this study would be the use of MEPS datasets that provide a careful design and execution of interviews, which involve multilevel verification of information collected from participants, such as checking with primary care physicians and insurance companies. The large sample size will increase the precision of the results to help draw inferences from the study to the American population.

Limitations to the 3-digit ICD-9-CM or ICD-10-CM code used to map medical conditions may make the perceived prevalence of ACS underestimated. Not all health providers are able to accurately match symptom presentation with the diagnosis of ACS, much less the appropriate ICD coding of that specific diagnosis. Second, since cardiovascular risk factors (LS7 components) for determining CVH scores were self-reported, the actual national prevalence is likely underestimated.

In addition, MEPS only conducts research among noninstitutionalized US civilians, and thus the results in this paper may represent the entire noninstitutionalized US population. Finally, the list of identified ICD codes for conditions classified as ACS might not be exhaustive because the authors focused on the most frequent conditions that define ACS.

Conclusion

A racial gap in CVH remains in place in the United States, and shows no sign of decline, despite how the adoption of the ACA likely has improved CVH management for both NH Whites and Blacks. This highlights how persistent racial disparities tend to be in the United States between Blacks and NH Whites. Blacks consistently have reduced odds of having ideal CVH and increased odds of having poor CVH, and until 2014, this bad situation was only improved if the trend away from healthy habits was even worse for NH Whites.

This research paper, which includes the impact of ACA, provides a comprehensive multi-year assessment of racial disparities in CVH management across this country. These findings may help motivate and refine efforts toward health equity from other sources, rather than ACA, or help to reform the ACA to target the racial gap. Efforts to create convergence in CVH to reduce disparities between Blacks and NH Whites in the United States would require government policies to look beyond mere “access” and/or “affordability” to health care.

Clinical Perspective

What is new?

• While the trend for improved CVH scores was on a downward trajectory across all 10 years, the advent of the ACA offset that trend by causing improvements in average scores by 3.6% per year across all 10 years.

• The racial gap in CVH scores remains in place in the United States, and shows no sign of decline, despite how the adoption of the ACA likely has improved CVH management for both NH Whites and Blacks. This highlights how persistent racial disparities tend to be in the United States between NH Blacks and Whites.

What are the clinical implications?

• The racial gap trend in CVH, though improved by the ACA, can lead to increased morbidity and mortality rates among minorities (NH Blacks) due to a lack of convergence. CVD is the leading cause of death for both NH Whites and Black Americans, but Blacks are at a higher risk of dying from it.

• CVD can have a significant impact on an individual's quality of life, and the racial gap in CVH can further exacerbate the negative effects of this disease on NH- Black communities.

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of Florida International University by the Office of Research Integrity (Research Compliance, MARC 414) on April 26, 2022. IRB-22-0170.

Author Disclosure Statement

The authors declare no conflict of interest.

Funding Information

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.