Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Feb 1.
Published in final edited form as: Ann Intern Med. 2023 Jul 25;176(8):1057–1066. doi: 10.7326/M23-0720

Disparities in Guideline-Recommended Statin Use for Prevention of Atherosclerotic Cardiovascular Disease by Race, Ethnicity, and Gender: A Nationally Representative Cross-Sectional Analysis of Adults in the United States

David A Frank 1,2, Amber E Johnson 3, Leslie R M Hausmann 1,4, Walid F Gellad 1,4, Eric T Roberts 5, Ravy K Vajravelu 1,6
PMCID: PMC10804313  NIHMSID: NIHMS1955659  PMID: 37487210

Abstract

Background:

Although statins are a Class I recommendation for prevention of atherosclerotic cardiovascular disease and its complications, their use is suboptimal. Differential underuse may mediate disparities in cardiovascular health for minoritized individuals.

Objective:

To estimate disparities in statin use by race-ethnicity-gender and to determine whether these potential disparities are explained medical appropriateness and structural factors.

Design:

Cross-sectional analysis

Setting:

National Health and Nutrition Examination Survey from 2015 – 2020

Participants:

Individuals eligible for statin therapy based on 2013 and 2018 ACC/AHA Blood Cholesterol Guidelines

Methods:

The independent variable was race-ethnicity-gender. The outcome of interest was use of a statin. Using the Institute of Medicine framework for examining unequal treatment, we calculated adjusted prevalence ratios (aPR) to estimate disparities in statin use adjusted for age, disease severity, access to healthcare, and socioeconomic status relative to non-Hispanic White men.

Results:

For primary prevention, we identified lower prevalence of statin use that was not explained by measurable differences in disease severity or structural factors among non-Hispanic Black men (aPR 0.73, 95% CI 0.59 – 0.88) and non-Mexican Hispanic women (aPR 0.74, 95% CI 0.53 – 0.95). For secondary prevention we identified lower prevalence of statin use that was not explained by measurable differences in disease severity or structural factors for non-Hispanic Black men (aPR 0.81, 95% CI 0.64 – 0.97), Other/Multiracial men (aPR 0.58, 95% CI 0.20 – 0.97), Mexican American women (aPR 0.36, 95% CI 0.10 – 0.61), non-Mexican Hispanic women (0.57, 95% CI 0.33 – 0.82), non-Hispanic White women (aPR 0.69, 95% CI 0.56 – 0.83), and non-Hispanic Black women (0.75, 95% CI 0.57 – 0.92).

Limitations:

Cross-sectional data; Lack of geographic, language, or statin-dose data

Conclusion:

Statin-use disparities for several race-ethnicity-gender groups are not explained by measurable differences in medical appropriateness of therapy, access to healthcare, and socioeconomic status. These residual disparities may be partially mediated by unobserved processes that contribute to health inequity, including bias, stereotyping, and mistrust.

Introduction

Preventing and treating atherosclerotic cardiovascular disease (ASCVD) has long been a major target of cardiovascular disease prevention programs (1). Statins, a class of medications that reduce low-density lipoprotein cholesterol (LDL-C) levels, are a mainstay of primary and secondary ASCVD prevention (2–5). However, rates of guideline-recommended statin use are suboptimal (6). The consequences of underutilization of statins may contribute to disparities in cardiovascular health outcomes. For example, non-Hispanic Black adults experience 32% higher age-adjusted rates of cardiovascular mortality compared to non-Hispanic White individuals (7–9). Similarly, cardiovascular disease is the leading cause of mortality among women, but advances in coronary heart disease management have imparted smaller mortality benefits to women compared to men (10–12).

Understanding racial, ethnic, and gender-based differences in statin use could inform strategies to improve population-level ASCVD outcomes. However, the current literature presents estimates of statin-use disparities derived from heterogeneous populations. A multi-institution registry of patients eligible for primary or secondary prevention of ASCVD demonstrated that Black individuals were four percent less likely to be using statins compared to White individuals (13). Similarly, among a sample of outpatient office visits in the United States, visits for non-Hispanic Black patients with ischemic heart disease had 25% lower odds of statin use for secondary prevention compared to visits for non-Hispanic White patients (14). Other studies have identified that women with coronary heart disease are less likely to use statins for secondary prevention compared to men with coronary heart disease (15, 16). While these studies establish the existence of statin use differences by race/ethnicity or gender, the interaction between race, ethnicity, and gender after adjustment for demographics, disease severity, and structural factors has not been studied for either primary or secondary prevention. These data are vital to identify strategies to mitigate cardiovascular disparities among minoritized individuals.

To address these knowledge gaps, we used nationally representative cross-sectional health status data from the National Health and Nutrition Examination Survey (NHANES) to estimate differences in guideline-recommended statin use by race, ethnicity, and gender for primary and secondary prevention of ASCVD. To assess whether these differences were explained by factors that influence the medical appropriateness of statin use, we calculated the prevalence of statin use adjusted for age and disease severity. Guided by the Institute of Medicine framework for evaluating unequal treatment, we interpreted these adjusted differences as disparities. Then, we additionally adjusted for measurable structural factors that influence access to healthcare, such as health insurance, formal education, and socioeconomic status. We interpreted differences in statin use that persisted after adjustment for measurable structural factors as residual disparities arising from unobserved care processes, including bias, stereotyping, and mistrust.

Methods

Additional details about the study methodology are presented in Supplemental methods.

Conceptual model and study approach

The conceptual model underlying the study approach was informed by the Institute of Medicine’s framework for evaluating unequal treatment (Figure 1) (17). This framework defines a disparity as a difference between groups that is not explained by patient factors that influence the appropriateness of a treatment. These disparities arise from socioeconomic and structural factors that influence access to healthcare, such as income and health insurance, and care-process factors that reflect a patient’s personal interactions with the healthcare system, including bias, discrimination, and mistrust.

Figure 1. Conceptual model depicting differences and disparities in statin use.

Figure 1.

Open in a new tab

The overall difference in statin use between the reference group and minoritized groups reflects patient factors that influence the medical appropriateness of statin use and disparities in access to statins. By accounting for patient factors such as age, disease severity, and medical comorbidities, disparities in statin use can be estimated. The influence of care-process factors, such as bias, stereotyping, and mistrust, can be estimated by also accounting for structural factors, such as health insurance and socioeconomic status indicators. Adapted from the Institute of Medicine framework for evaluating unequal treatment (16).

Data source

NHANES is a repeated cross-sectional study conducted by the Centers for Disease Control and Prevention to assess the health and nutrition status of individuals living in the United States through interviews, physical examination, and biometric measurements. In each cycle, subjects are selected to participate through multistage, stratified, random sampling, yielding a study population of approximately 10,000 individuals. Survey weights are provided to generate nationally representative estimates. To assess statin-use disparities in reference to 2013 and 2018 American College of Cardiology/American Heart Association (ACC/AHA) blood cholesterol guidelines, we utilized NHANES data from the 2015 – 2016 and 2017 – pre-pandemic 2020 cycles (18, 19). NHANES protocols and data analysis have been approved by the National Center for Health Statistics Ethics Review Board (20).

Population of interest: Individuals eligible for statin use

We focused on populations advised to receive statin therapy by Class I recommendations in the 2013 and 2018 ACC/AHA guidelines (1, 2). For primary prevention this included individuals without ASCVD who were ages 21 – 75 with LDL-C ≥190 mg/dL, ages 40 – 75 with diabetes, or ages 40 – 75 with 10-year ASCVD risk ≥7.5%. For secondary prevention, this included individuals ages 21 – 75 with clinical ASCVD. Per ACC/AHA guidelines, we defined ASCVD as history of acute coronary syndrome, history of myocardial infarction, stable or unstable angina, coronary or other arterial revascularization, stroke, transient ischemic attack, or peripheral arterial disease. We identified these diagnoses using data from the NHANES medical questionnaire, in which participants report whether they have conditions of interest based on structured interviews. We ascertained LDL-C using the NHANES fasting examination data. We calculated ten-year ASCVD risk using data from questionnaires, the general examination, and the fasting examination according to the 2013 ACC/AHA ASCVD risk calculator (21).

Outcome of interest: Statin use

The main outcome of interest was statin use among individuals eligible to receive statins. This was ascertained from the prescription medications questionnaire. To determine if a participant is using a prescription medication, NHANES staff interview participants in their homes. Medications taken within the prior 30 days are reported by the participant. The interviewer records the name of the medication by matching the product to a prescription drug database. Reported prescriptions are also verified by comparison to pill bottles when available.

Independent variable: Race-ethnicity-gender category

In NHANES, race and ethnicity are self-identified by participants and reported as a single variable categorized as Mexican American, Other Hispanic (non-Mexican Hispanic henceforth), non-Hispanic White, non-Hispanic Black, non-Hispanic Asian, and Other/Multiracial. Publicly available NHANES data do not report race-ethnicity for Native American/Alaska Native, or Pacific Islander individuals (22). We refer to this variable as race-ethnicity to denote that it reflects elements of both race and ethnicity. Because there are also established health disparities by gender, we stratified race-ethnicity category by gender, thereby reflecting the interaction between these variables (16). Transgender and non-binary gender identities are not reported in NHANES. All combinations of race-ethnicity-gender were compared to the reference category of non-Hispanic White men. We recognize that any group could have served as the referent, but we chose non-Hispanic White men because we hypothesized that this group would have the highest rates of statin use, thereby highlighting potential disparities (23–26).

Explanatory factors for differences and disparities in statin use

Guided by the Institute of Medicine framework for examining unequal treatment, we categorized covariates into two groups (17): First, we identified patient factors potentially related to the medical appropriateness of statin therapy. Second, we identified structural factors related to access to healthcare and socioeconomic status. The individual covariates evaluated are listed in Figure 1. Additional details about how these covariates were categorized and ascertained are presented in Supplemental methods. In other contexts, these explanatory factors could be considered moderators or mediators of statin use by race-ethnicity-gender. However, because of the cross-sectional study design, we were not able to ascertain the temporality between the factors and statin use, so we refer to them as “explanatory factors” instead of “moderators/mediators.” Missing covariates were imputed ten times via the chained equations method using data from the entire NHANES cohort from 2015 – pre-pandemic 2020 (the percentage of missing observations by variable are presented in Supplemental table 1).

Statistical analysis

To assess the extent that the explanatory factors described above contribute to differences in the prevalence of statin use by race-ethnicity-gender, we developed three logistic regression models. Specification of these models was guided by the Institute of Medicine framework for evaluating unequal treatment. Each model calculated the survey-weighted log odds of statin use by race-ethnicity-gender category. We then calculated the marginally adjusted prevalence of statin use for each race-ethnicity-gender group using the mimrgns command from Stata among the relevant subpopulation (27). Finally, using the nlcom command, we compared the marginally adjusted prevalence of statin use for each race-ethnicity-gender group to the marginally adjusted prevalence of statin use for non-Hispanic White men (reference group) to obtain adjusted prevalence ratios (aPRs) with 95% confidence intervals (CIs).

Model 1 was unadjusted for the explanatory factors described above and estimated the difference in statin use by race-ethnicity-gender. Model 2 was adjusted for measurable patient factors related to the medical appropriateness of statin therapy. We interpreted lower prevalence of statin use in model 2 as evidence of a disparity in statin use by race-ethnicity-gender. Model 3 adjusted for measurable patient factors and structural factors. We interpreted lower prevalence of statin use in model 3 as a residual disparity reflecting unobserved processes that influence health inequity, including bias, stereotyping, and mistrust (17). Each of the three models were evaluated separately for primary prevention of ASCVD and secondary prevention of ASCVD complications.

Because NHANES is a probabilistic survey, all estimates were weighted using primary sampling units and masked variance pseudo-strata to generalize the results to the United States population. Analyses were conducted using appropriate svy and subpop commands in Stata. Analyses utilizing laboratory data were weighted by the fasting subsample weight instead of the interview weight, and weights were scaled to account for combining data from the 2015 – 2016 and 2017 – pre-pandemic 2020 cycles. Data extraction, cleaning, and analyses were performed using Stata/SE 17.0 (College Station, Texas). Figure 1 was generated using BioRender. Figures 2 and 3 were generated using ggplot2 package in R (Vienna, Austria) (28, 29).

Figure 2. aPRs for statin use for primary prevention of ASCVD by race-ethnicity-gender category.

Figure 2.

Open in a new tab

Model 1 was unadjusted, thereby reflecting differences in statin use that are influenced by patient factors that contribute to the medical appropriateness of care, structural factors, and care-process factors. Model 2 was adjusted for measurable patient factors only, thereby reflecting disparities that are influenced by structural factors and care-process factors. Model 3 was adjusted for measurable patient factors and structural factors, thereby reflecting disparities unexplained by structural factors. These residual disparities may reflect unobserved mediators of health inequity, including bias, stereotyping, and mistrust. Results weighted to the United States population. Adjusted PRs were derived from marginally adjusted prevalence estimates calculated from multivariable logistic regression models. Adjusted PRs for covariates are presented in Supplemental table 3.

Figure 3. aPRs for statin use for secondary prevention of ASCVD complications by race-ethnicity-gender category.

Figure 3.

Open in a new tab

Model 1 was unadjusted, thereby reflecting differences in statin use that are influenced by patient factors that contribute to the medical appropriateness of care, structural factors, and care-process factors. Model 2 was adjusted for measurable patient factors only, thereby reflecting disparities that are influenced by structural factors and care-process factors. Model 3 was adjusted for measurable patient factors and structural factors, thereby reflecting disparities unexplained by structural factors. These residual disparities may reflect unobserved mediators of health inequity, including bias, stereotyping, and mistrust. Results weighted to the United States population. Adjusted PRs were derived from marginally adjusted prevalence estimates calculated from multivariable logistic regression models. Adjusted PRs for covariates are presented in Supplemental table 4.

Sensitivity analyses

We conducted four sensitivity analyses to test our methodologic assumptions. First, because recent studies suggest that the 2013 ACC/AHA ASCVD risk calculator may overestimate risk for Black individuals (30, 31), we performed a sensitivity analysis to assess whether defining ASCVD risk without accounting for race changes conclusions about residual disparities. Second, because there were other primary prevention guidelines released after 2013 with higher ASCVD risk thresholds, we performed a sensitivity analysis to assess residual disparities when the ASCVD risk threshold was set at 20% instead of 7.5% (32). Third, because NHANES is a cross-sectional study, it is possible that some participants using statins were misclassified as not being eligible for primary prevention due to improvement of LDL-C and ASCVD risk. To assess the impact of this misclassification, we performed a sensitivity analysis in which all individuals treated with a statin and not eligible for secondary prevention were classified as eligible for primary prevention. Fourth, because angina as an indication for secondary prevention could be misclassified by recall bias, we conducted a sensitivity analysis in which participants who reported angina as their only indication for secondary prevention were excluded.

Role of the funding source

This research was funded by the National Institutes of Health. The funders of this study had no role in the study design, data collection, data analysis, data interpretation, or presentation of the results.

Results

We identified 13,213 participants in NHANES ages 21 to 75, representing 216,854,504 individuals in the United States population. Four thousand seven hundred sixty-three participants, representing 62,290,942 individuals, were eligible to receive a statin for primary prevention of ASCVD. The indications for primary prevention were one or more of LDL-C ≥ 190 mg/dL (7.5%), diabetes (25.8%), or 10-year ASCVD risk ≥ 7.5% (87.6%). One thousand one hundred thirty-eight participants, representing 16,548,722 individuals, were eligible to receive a statin for secondary prevention of ASCVD complications. The indications for secondary prevention were one or more of history of angina (25.5%), coronary heart disease (45.2%), myocardial infarction (41.8%), or stroke (38.1%). Additional characteristics of the study populations are presented in Table 1.

Table 1.

Characteristics of NHANES and statin-eligible participants

Primary prevention (unweighted n = 4,763 weighted n = 62,290,942) Secondary prevention (unweighted n = 1,138 weighted n = 16,548,722)
Race-ethnicity (%)
Men
 Mexican American 3.9 (2.9 – 5.1) 2.3 (1.5 – 3.5)
 Non-Mexican Hispanic 3.6 (2.8 – 4.8) 2.5 (1.6 – 3.9)
 Non-Hispanic White 38.2 (34.2 – 42.3) 40.2 (33.2 – 47.6)
 Non-Hispanic Black 5.8 (4.5 – 7.4) 5.8 (3.9 – 8.6)
 Non-Hispanic Asian 3.0 (2.2 – 4.2) 1.9 (1.0 – 3.5)
 Other/Multiracial 1.9 (1.1 – 3.1) 5.3 (2.7 – 10.0)
Women
 Mexican American 3.0 (2.0 – 4.4) 1.3 (0.7 – 2.2)
 Non-Mexican Hispanic 2.7 (1.9 – 3.7) 3.5 (2.3 – 5.4)
 Non-Hispanic White 28.3 (24.5 – 32.5) 25.0 (18.3 – 33.0)
 Non-Hispanic Black 6.0 (4.5 – 8.0) 7.7 (5.7 – 10.5)
 Non-Hispanic Asian 2.2 (1.6 – 3.0) 0.8 (0.4 – 1.7)
 Other/Multiracial 1.6 (1.0 – 2.7) 3.8 (1.9 – 7.2)
Patient factors (% unless otherwise specified)
Age (mean) 62.0 (61.1 – 62.9) 59.8 (58.5 – 61.2)
10-year ASCVD risk (mean) 17.8 (17.0 – 18.6) –
History of coronary heart disease – 41.2 (34.0 – 48.4)
History of myocardial infarction – 41.5 (32.0 – 51.1)
History of stroke – 37.3 (31.3 – 43.2)
Family history of myocardial infarction 16.1 (13.7 – 18.8) 33.6 (25.5 – 41.5)
Use of fibrates 1.8 (1.1 – 2.4) 3.5 (1.4 – 5.7)
Body mass index (mean) 30.3 (29.9 – 30.7) 31.2 (30.3 – 32.1)
History of cancer 18.5 (16.1 – 20.9) 20.1 (14.3 – 25.9)
History of chronic kidney disease 4.5 (3.0 – 6.0) 7.5 (4.0 – 10.9)
Estimated glomerular filtration rate (mean) 84.8 (83.1 – 86.5) 82.6 (79.7 – 85.4)
History of chronic obstructive pulmonary disease 18.7 (15.5 – 21.9) 34.4 (25.6 – 43.2)
History of diabetes 29.2 (26.2 – 32.3) 32.8 (25.8 – 39.9)
Glycosylated hemoglobin (mean) 6.2 (6.1 – 6.3) 6.3 (6.1 – 6.5)
History of heart failure 1.9 (1.4 – 2.4) 17.9 (12.3 – 23.5)
History of liver disease 7.1 (4.8 – 9.3) 9.4 (5.6 – 13.1)
Number of medications (mean) 3.3 (3.1 – 3.6) 5.5 (4.9 – 6.0)
Number of statin intolerance risk factors* (mean) 0.6 (0.5 – 0.7) 0.8 (0.7 – 1.0)
Health care visits last year (mean) 4.5 (4.2 – 4.8) 6.1 (5.6 – 6.7)
Hospitalization last year 11.0 (8.8 – 13.7) 25.9 (21.6 – 30.7)
Number of ready-to-eat meals in the last 30 days (mean) 2.2 (1.8 – 2.6) 2.3 (1.5 – 3.0)
Moderate or greater physical activity regularly 68.7 (65.5 – 71.8) 67.8 (62.6 – 72.7)
Self-perceived health
 Poor 3.2 (2.4 – 4.4) 12.2 (8.0 – 18.0)
 Fair 18.8 (16.2 – 21.6) 30.9 (24.7 – 37.8)
 Good 39.7 (37.4 – 42.2) 35.8 (30.6 – 41.4)
 Very good 26.8 (23.8 – 30.1) 18.5 (11.2 – 28.9)
 Excellent 11.5 (9.6 – 13.6) 2.7 (1.4 – 5.0)
Structural factors (%)
Health insurance
 None 7.4 (5.8 – 9.4) 12.2 (8.2 – 17.7)
 Private 31.1 (27.7 – 34.7) 23.6 (18.1 – 30.2)
 Medicare 10.7 (8.6 – 13.2) 14.5 (9.7 – 21.0)
 Medicaid 3.7 (2.8 – 4.9) 5.9 (3.5 – 9.7)
 Other government 6.3 (5.0 – 7.9) 7.9 (5.0 – 12.2)
 More than one type 40.5 (36.9 – 44.1) 35.5 (29.1 – 42.5)
 Unspecified 0.4 (0.2 – 0.8) 0.5 (0.2 – 1.4)
Prescription medication coverage 92.8 (90.5 – 94.5) 91.7 (87.1 – 94.8)
Formal education level
 Less than high school 14.7 (12.3 – 17.5) 19.1 (14.0 – 25.6)
 High school 56.4 (51.7 – 60.9) 59.4 (52.3 – 66.2)
 College or more 28.9 (25.0 – 33.2) 21.5 (14.3 – 30.9)
Household income relative to federal poverty level
 Less than or equal to 1.30 21.3 (18.0 – 25.1) 28.6 (21.9 – 36.4)
 1.30 to 1.85 11.5 (9.6 – 13.7) 15.5 (11.8 – 20.2)
 > 1.85 67.2 (62.9 – 71.3) 55.8 (47.6 – 63.8)
Partnered 67.0 (63.0 – 70.8) 64.7 (57.6 – 71.2)
Regular place for health care 92.0 (90.2 – 93.5) 92.9 (89.1 – 95.5)
Open in a new tab

Notes: Results weighted to the United States population.

*

Statin intolerance risk factors: Current liver disease, current hypothyroidism, chronic kidney disease, heavy alcohol use, use of a medication with a known pharmacokinetic interaction with any statin.

Prevalence of statin use for primary prevention by race-ethnicity-gender

The overall prevalence of statin use for primary prevention was 37.6% (95% CI 33.9 – 41.5). Unadjusted prevalence of statin use for primary prevention ranged from 23.8% (95% CI 18.5 – 30.0) for non-Hispanic Black men to 41.1% (95% CI 33.5 – 49.1) for non-Hispanic White women (Supplemental table 2). Model 1, which was unadjusted for explanatory factors, demonstrated lower prevalence of statin use among non-Hispanic Black men (aPR 0.61, 95% CI 0.45 – 0.76) relative to non-Hispanic White men (Figure 2, Supplemental table 3). Model 2, which assessed differences in the prevalence of statin use adjusted for measurable patient factors that influence the medical appropriateness of therapy, demonstrated lower prevalence of statin use among non-Hispanic Black men (aPR 0.70, 95% CI 0.55 – 0.84), non-Mexican Hispanic women (aPR 0.70, 95% CI 0.51 – 0.89), non-Hispanic White women (aPR 0.82, 95% CI 0.66 – 0.99), non-Hispanic Black women (aPR 0.84, 95% CI 0.69 – 0.99), and non-Hispanic Asian women (aPR 0.73, 95% CI 0.51 – 0.94) relative to non-Hispanic White men. Model 3, which assessed residual disparities not explained by measurable differences in medical appropriateness of therapy, access to healthcare, and socioeconomic status, demonstrated lower prevalence of statin use for non-Hispanic Black men (aPR 0.73, 95% CI 0.59 – 0.88) and non-Mexican Hispanic women (aPR 0.74, 95% CI 0.53 – 0.95) relative to non-Hispanic White men.

Prevalence of statin use for secondary prevention by race-ethnicity-gender

The overall prevalence of statin use for secondary prevention was 59.1% (95% CI 54.8 – 63.2). Unadjusted prevalence of statin use for secondary prevention ranged from 37.6% (95% CI 25.1 – 52.1) for non-Mexican Hispanic women to 71.8% (95% CI 64.5–78.0) for non-Hispanic White men (Supplemental table 2). Model 1, which was unadjusted for explanatory factors, demonstrated lower prevalence of statin use among Mexican American men (aPR 0.71, 95% CI 0.49 – 0.93), non-Hispanic Black men (aPR 0.66, 95% CI 0.56 – 0.76), Mexican American women (aPR 0.57, 95% CI 0.29 – 0.86), non-Mexican Hispanic women (aPR 0.52, 95% CI 0.34 – 0.71), non-Hispanic White women (aPR 0.68, 95% CI 0.56 – 0.79), and non-Hispanic Black women (aPR 0.53, 95% CI 0.40 – 0.66) relative to non-Hispanic White men (Figure 3, Supplemental table 4). Model 2, which assessed differences in the prevalence of statin use adjusted for measurable patient factors that influence the medical appropriateness of therapy, demonstrated lower prevalence of statin use among Mexican American women (aPR 0.44, 95% CI 0.22 – 0.67), non-Mexican Hispanic women (aPR 0.59, 95% CI 0.32 – 0.85), non-Hispanic White women (0.73, 95% CI 0.57 – 0.90), and non-Hispanic Black women (aPR 0.76, 95% CI 0.57 – 0.95) relative to non-Hispanic White men. Model 3, which assessed residual disparities not explained by measurable differences in medical appropriateness of therapy, access to healthcare, and socioeconomic status, demonstrated lower prevalence of statin use for non-Hispanic Black men (aPR 0.81, 95% CI 0.64 – 0.97), Other/Multiracial men (aPR 0.58, 95% CI 0.20 – 0.97), Mexican American women (aPR 0.36, 95% CI 0.10 – 0.61), non-Mexican Hispanic women (aPR 0.57, 95% CI 0.33 – 0.82), non-Hispanic White women (0.69, 95% CI 0.56 – 0.83), and non-Hispanic Black women (aPR 0.75, 95% CI 0.57 – 0.92).

Sensitivity analyses

In the sensitivity analysis assessing the impact of excluding race from the ASCVD risk calculator, conclusions about residual disparities were unchanged. In the sensitivity analysis assessing the impact of increasing the ASCVD threshold to 20%, conclusions about residual disparities were unchanged for non-Hispanic Black men, but the residual disparity for non-Mexican Hispanic women was attenuated (APR 0.88, 95% CI 0.65 – 1.11). In the sensitivity analysis assessing the impact of expanding the primary prevention eligibility to include statin-treated individuals who did not meet other criteria, conclusions about residual disparities were unchanged for non-Hispanic Black men, but the residual disparity for non-Mexican Hispanic women was attenuated (aPR 0.86, 95% CI 0.64 – 1.07). Full results are presented in Supplemental table 5. For the sensitivity analysis assessing the potential impact of misclassification of secondary prevention eligibility by recall bias for angina, 8% of participants eligible for secondary prevention in the main analysis were excluded. Conclusions about residual disparities were unchanged from the main analysis (Supplemental table 6).

Discussion

In this cross-sectional study using data from NHANES from 2015 – 2020, we utilized the Institute of Medicine framework for evaluation of unequal treatment to identify disparities in the prevalence of statin use among a nationally representative group of individuals eligible for primary prevention of ASCVD or secondary prevention of ASCVD complications. By adjusting for measurable patient and structural factors that reflect disease severity and unequal access to resources, we found lower prevalence of statin use for primary prevention among non-Hispanic Black men and non-Mexican Hispanic women compared to non-Hispanic White men. For secondary prevention, we found lower prevalence of statin use for non-Hispanic Black men, Other/Multiracial men, Mexican American women, non-Mexican Hispanic women, non-Hispanic White women, and non-Hispanic Black women compared to non-Hispanic White men. As these residual disparities are not explained by measurable disease severity factors, such as cardiovascular comorbidities, or measurable differences in healthcare or socioeconomic resources, such as health insurance, formal education, and income, they may partially reflect the role of unobserved factors that influence health inequities, including bias, stereotyping, and mistrust.

These findings build on prior literature identifying cardiovascular disease prevention disparities that are not explained by measurable differences in socioeconomic factors. For example, Johnson et al. recently demonstrated that higher levels of educational attainment, often considered a mediator of health over the life course, has less impact on cardiometabolic health for minoritized individuals compared to non-Hispanic White individuals (33). This result signifies that economic opportunity is not the sole contributor to cardiovascular health disparities. This conclusion was underscored by a recent study by O’Hearn et al. that demonstrated that the percentage of individuals with optimal cardiometabolic health in the United States increased among non-Hispanic White individuals from 1999 to 2018, but declined for all other races and ethnicities (24). Our work also adds to the literature showing disparities in ASCVD management beyond statin use. For example, Tertulian et al. used medical claims data from commercial and Medicare Advantage health plans in the United States to demonstrate that Black and Hispanic individuals are less likely to receive left heart catheterization or percutaneous coronary intervention after non-ST-segment elevation myocardial infarction compared to White individuals, even after accounting for income and education (25). Similarly, Garfein et al. demonstrated that Asian, Black, and Hispanic individuals eligible for cardiac rehabilitation were less likely to participate compared to White individuals, regardless of income level (26). Collectively, these studies demonstrate the presence of disparities in care after adjustment for structural factors linked to socioeconomic status and education, thereby suggesting that factors within the health care setting, including bias, stereotyping, and mistrust, contribute to unequal treatment.

Our analysis has several strengths that lend credibility to the study conclusions. First, the data are nationally representative, reflecting health care experiences across the United States for individuals from several race-ethnicity-gender categories. Second, unlike many sources of administrative healthcare data where race and ethnicity data are statistically imputed based on naming and geographic patterns, race-ethnicity data in NHANES are self-reported, thereby minimizing misclassification and reflecting each participant’s lived experience interacting with societal and health institutions. Third, we used a rich set of demographic, health, access to care, and socioeconomic variables in NHANES to adjust for the domains of disease severity, medical comorbidities, health system interactions, and socioeconomic status. Finally, the residual disparity in statin use for primary prevention among non-Hispanic Black men was demonstrated consistently across three sensitivity analyses.

This study also has limitations that must be considered when interpreting the results. First, because of the length of the NHANES survey cycles, we were not able to study whether the disparities in guideline-recommended statin use by race-ethnicity-gender diminished over time. The first relevant ACC/AHA blood cholesterol guideline for this study was released within the 2013 – 2014 NHANES cycle. To allow adequate time for diffusion of the guideline, we started this study with the 2015 – 2016 NHANES cycle. Because of the COVID-19 pandemic, the 2019 – 2020 cycle was truncated and combined with the 2017 – 2018 cycle. Therefore, we were only able to study two cycles of NHANES data, making presentation of trends infeasible. Second, although we adjusted for several explanatory factors to account for differences and disparities in statin use in each analysis, we cannot definitively exclude unmeasured confounding. There may be other factors, such as neighborhood built environment, access to transportation, and English proficiency, that mediate health disparities but are not available in NHANES (34–39). Third, because NHANES does not contain medication dose information, we were not able to account for statin intensity. Because the ACC/AHA guidelines recommended high-intensity statins for some risk groups, accounting for statin dose in future analyses may identify even larger health disparities in guideline-concordant care. Fourth, because NHANES suppresses geographic data for participant privacy, we were not able to assess sociopolitical mediators of residual disparities among the race-ethnicity-gender groups. Finally, due to the cross-sectional nature of NHANES, we could not determine whether statins had been previously prescribed for the individuals recommended for statins but not currently receiving them. Availability of those data could help distinguish what portion of the disparity not explained by patient or structural factors is related to bias and stereotyping emanating from the prescriber versus the portion that is related to mistrust from the patient towards the provider or healthcare system (40).

In conclusion, this study demonstrates disparities in the prevalence of statin use for primary prevention of ASCVD among non-Hispanic Black men and Non-Mexican Hispanic women and in the prevalence of statin use for secondary prevention of ASCVD complications among non-Hispanic Black men, Other/Multiracial men, non-Mexican Hispanic women, non-Hispanic White women, and non-Hispanic Black women. These disparities were not explained by measurable differences in disease severity or access to resources, and therefore may be partially mediated by care-process factors, including bias, stereotyping, and mistrust. Because these statin use disparities may contribute to disparities in overall cardiovascular morbidity and mortality, they highlight the importance of societal interventions to health delivery systems to reduce inequity in care delivery and treatment (7–12). These include clinical quality improvement initiatives to systematize statin prescriptions among eligible patients, bias reduction training for prescribers, diversification of the health care provider workforce, and programs to regain trust among minoritized groups who have experienced intergenerational scientific and clinical misconduct (41, 42).

Supplementary Material

Supplementary material

Funding source:

National Institutes of Health

Grant Support:

  • Ravy K. Vajravelu: NIH/NIDDK—K08-DK119475

Abbreviations

ACC/AHA

American College of Cardiology/American Heart Association

aPR

Adjusted prevalence ratio

ASCVD

Atherosclerotic cardiovascular disease

CI

Confidence interval

LDL-C

Low-density lipoprotein cholesterol

NHANES

National Health and Nutrition Examination Survey

Footnotes

Disclosures:

All authors are employees of the Department of Veterans Affairs. This research does not represent the views of Department of Veterans Affairs or the United States Government.

Data sharing:

Upon publication of the manuscript, computer code used to generate the study results will be posted to the corresponding author’s GitHub repository at https://github.com/rvajravelu with a README file describing the function of each file. These files will be freely available. Study protocol: Study protocols will not be made available. Statistical code: Computer code used to generate the study results are freely available at https://github.com/rvajravelu/statinDisparities with a README file describing implementation instructions. Data set: All raw data are freely available via the National Center for Health Statistics at www.cdc.gov/nchs/nhanes/index.htm.

REFERENCES

  • 1.Goodman DS, Hulley SB, Clark LT, David CE, Fuster V, LaRosa JC, et al. Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. The Expert Panel. Arch Intern Med. 1988;148(1):36–69. [PubMed] [Google Scholar]
  • 2.Stone NJ, Robinson JG, Lichtenstein AH, Bairey Merz CN, Blum CB, Eckel RH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2014;63(25 Pt B):2889–934. [DOI] [PubMed] [Google Scholar]
  • 3.Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;139(25):e1082–e143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Mangione CM, Barry MJ, Nicholson WK, Cabana M, Chelmow D, Coker TR, et al. Statin Use for the Primary Prevention of Cardiovascular Disease in Adults: US Preventive Services Task Force Recommendation Statement. JAMA. 2022;328(8):746–53. [DOI] [PubMed] [Google Scholar]
  • 5.Mihaylova B, Emberson J, Blackwell L, Keech A, Simes J, Barnes EH, et al. The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials. Lancet. 2012;380(9841):581–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Saeed A, Zhu J, Thoma F, Marroquin O, Aiyer A, Agarwala A, et al. Cardiovascular Disease Risk-Based Statin Utilization and Associated Outcomes in a Primary Prevention Cohort: Insights From a Large Health Care Network. Circ Cardiovasc Qual Outcomes. 2021;14(9):e007485. [DOI] [PubMed] [Google Scholar]
  • 7.Mahajan S, Caraballo C, Lu Y, Valero-Elizondo J, Massey D, Annapureddy AR, et al. Trends in Differences in Health Status and Health Care Access and Affordability by Race and Ethnicity in the United States, 1999–2018. JAMA. 2021;326(7):637–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Murphy S, Kochanek K, Xu J, Arias E. Mortality in the United States, 2020. NCHS Data Brief, no 427. Hyattsville, MD: National Center for Health Statistics; 2021. [Google Scholar]
  • 9.Data Visualizations. Accessed at American Heart Association Center for Health Metrics and Evaluation (CHIME) at https://www.heart.org/en/about-us/2024-health-equity-impact-goal/age-adjusted-total-cvd-mortality-rates-by-race-ethnicity on March 15, 2023. [Google Scholar]
  • 10.Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Carson AP, et al. Heart Disease and Stroke Statistics-2019 Update: A Report From the American Heart Association. Circulation. 2019;139(10):e56–e528. [DOI] [PubMed] [Google Scholar]
  • 11.Wilmot KA, O’Flaherty M, Capewell S, Ford ES, Vaccarino V. Coronary Heart Disease Mortality Declines in the United States From 1979 Through 2011: Evidence for Stagnation in Young Adults, Especially Women. Circulation. 2015;132(11):997–1002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.El Khoudary SR, Aggarwal B, Beckie TM, Hodis HN, Johnson AE, Langer RD, et al. Menopause Transition and Cardiovascular Disease Risk: Implications for Timing of Early Prevention: A Scientific Statement From the American Heart Association. Circulation. 2020;142(25):e506–e32. [DOI] [PubMed] [Google Scholar]
  • 13.Nanna MG, Navar AM, Zakroysky P, Xiang Q, Goldberg AC, Robinson J, et al. Association of Patient Perceptions of Cardiovascular Risk and Beliefs on Statin Drugs With Racial Differences in Statin Use: Insights From the Patient and Provider Assessment of Lipid Management Registry. JAMA Cardiol. 2018;3(8):739–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Minhas AMK, Ijaz SH, Javed N, Sheikh AB, Jain V, Michos ED, et al. National trends and disparities in statin use for ischemic heart disease from 2006 to 2018: Insights from National Ambulatory Medical Care Survey. Am Heart J. 2022;252:60–9. [DOI] [PubMed] [Google Scholar]
  • 15.Johansen ME, Hefner JL, Foraker RE. Antiplatelet and Statin Use in US Patients With Coronary Artery Disease Categorized by Race/Ethnicity and Gender, 2003 to 2012. Am J Cardiol. 2015;115(11):1507–12. [DOI] [PubMed] [Google Scholar]
  • 16.Shah T, Palaskas N, Ahmed A. An Update on Gender Disparities in Coronary Heart Disease Care. Curr Atheroscler Rep. 2016;18(5):28. [DOI] [PubMed] [Google Scholar]
  • 17.Introduction and Literature Review. In: Smedley BD, Stith AY, Nelson AR, eds. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington (DC); 2003:29–77. [PubMed] [Google Scholar]
  • 18.NHANES 2015–2016 Overview. Centers for Disease Control and Prevention; at https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/overview.aspx?BeginYear=2015 on October 3, 2022. [Google Scholar]
  • 19.NHANES Analytic Guidance and Brief Overview for the 2017-March 2020 Pre-pandemic Data Files. Centers for Disease Control and Prevention; at https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/overviewbrief.aspx?Cycle=2017-2020 on October 3, 2022. [Google Scholar]
  • 20.About the National Health and Nutrition Examination Survey. Centers for Disease Control and Prevention; at https://www.cdc.gov/nchs/nhanes/about_nhanes.htm on October 3, 2022. [Google Scholar]
  • 21.Goff DC Jr., Lloyd-Jones DM, Bennett G, Coady S, D’Agostino RB, Gibbons R, et al. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014;129(25 Suppl 2):S49–73. [DOI] [PubMed] [Google Scholar]
  • 22.Ponce N, Becker T, Babey S, Shimkhada R, Hanaya D, Lee S, et al. Improving Data Capacity for American Indian/Alaska Native (AIAN) Populations in Federal Health Surveys. Office of the Assitatnt Secretary for Planning and Evaluation, Department of Health and Human Services; 2019. [Google Scholar]
  • 23.Johfre SS, Freese J. Reconsidering the Reference Category. Sociological Methodology. 2021;51(2):253–69. [Google Scholar]
  • 24.O’Hearn M, Lauren BN, Wong JB, Kim DD, Mozaffarian D. Trends and Disparities in Cardiometabolic Health Among U.S. Adults, 1999–2018. J Am Coll Cardiol. 2022;80(2):138–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Tertulien T, Broughton ST, Swabe G, Essien UR, Magnani JW. Association of Race and Ethnicity on the Management of Acute Non-ST-Segment Elevation Myocardial Infarction. J Am Heart Assoc. 2022;11(12):e025758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Garfein J, Guhl EN, Swabe G, Sekikawa A, Barinas-Mitchell E, Forman DE, et al. Racial and Ethnic Differences in Cardiac Rehabilitation Participation: Effect Modification by Household Income. J Am Heart Assoc. 2022;11(13):e025591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Klein D. MIMRGNS: Stata module to run margins after mi estimate. 2014.
  • 28.H W. ggplot2: Elegant Graphics for Data Analysis. 2016. 978-3-319-24277-4. https://ggplot2.tidyverse.org.
  • 29.Team RC. R: A language and environment for statistical computing. R Foundation for Statistical Computing. 2022. https://www.R-project.org. [Google Scholar]
  • 30.Vasan RS, van den Heuvel E. Differences in estimates for 10-year risk of cardiovascular disease in Black versus White individuals with identical risk factor profiles using pooled cohort equations: an in silico cohort study. Lancet Digit Health. 2022;4(1):e55–e63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Vyas DA, James A, Kormos W, Essien UR. Revising the atherosclerotic cardiovascular disease calculator without race. Lancet Digit Health. 2022;4(1):e4–e5. [DOI] [PubMed] [Google Scholar]
  • 32.Bibbins-Domingo K, Grossman DC, Curry SJ, Davidson KW, Epling JW Jr., Garcia FAR, et al. Statin Use for the Primary Prevention of Cardiovascular Disease in Adults: US Preventive Services Task Force Recommendation Statement. JAMA. 2016;316(19):1997–2007. [DOI] [PubMed] [Google Scholar]
  • 33.Johnson AE, Herbert BM, Stokes N, Brooks MM, Needham BL, Magnani JW. Educational Attainment, Race, and Ethnicity as Predictors for Ideal Cardiovascular Health: From the National Health and Nutrition Examination Survey. J Am Heart Assoc. 2022;11(2):e023438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Johnson AE, Zhu J, Garrard W, Thoma FW, Mulukutla S, Kershaw KN, et al. Area Deprivation Index and Cardiac Readmissions: Evaluating Risk-Prediction in an Electronic Health Record. J Am Heart Assoc. 2021;10(13):e020466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Guadamuz JS, Wilder JR, Mouslim MC, Zenk SN, Alexander GC, Qato DM. Fewer Pharmacies In Black And Hispanic/Latino Neighborhoods Compared With White Or Diverse Neighborhoods, 2007–15. Health Aff (Millwood). 2021;40(5):802–11. [DOI] [PubMed] [Google Scholar]
  • 36.Auchincloss AH, Diez Roux AV, Mujahid MS, Shen M, Bertoni AG, Carnethon MR. Neighborhood resources for physical activity and healthy foods and incidence of type 2 diabetes mellitus: the Multi-Ethnic study of Atherosclerosis. Arch Intern Med. 2009;169(18):1698–704. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Auchincloss AH, Mujahid MS, Shen M, Michos ED, Whitt-Glover MC, Diez Roux AV. Neighborhood health-promoting resources and obesity risk (the multi-ethnic study of atherosclerosis). Obesity (Silver Spring). 2013;21(3):621–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Osypuk TL, Diez Roux AV, Hadley C, Kandula NR. Are immigrant enclaves healthy places to live? The Multi-ethnic Study of Atherosclerosis. Soc Sci Med. 2009;69(1):110–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Essien UR, Dusetzina SB, Gellad WF. A Policy Prescription for Reducing Health Disparities-Achieving Pharmacoequity. JAMA. 2021;326(18):1793–4. [DOI] [PubMed] [Google Scholar]
  • 40.Brown CJ, Chang LS, Hosomura N, Malmasi S, Morrison F, Shubina M, et al. Assessment of Sex Disparities in Nonacceptance of Statin Therapy and Low-Density Lipoprotein Cholesterol Levels Among Patients at High Cardiovascular Risk. JAMA Netw Open. 2023;6(2):e231047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.The U.S. Public Health Service Syphilis Study at Tuskegee. Centers for Disease Control and Prevention; at https://www.cdc.gov/tuskegee/index.html on February 21, 2023. [Google Scholar]
  • 42.Lawrence J. The Indian Health Service and the Sterilization of Native American Women. American Indian Quarterly. 2000;24:400–19. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary material
NIHMS1955659-supplement-Supplementary_material.pdf (271.8KB, pdf)

Data Availability Statement

Upon publication of the manuscript, computer code used to generate the study results will be posted to the corresponding author’s GitHub repository at https://github.com/rvajravelu with a README file describing the function of each file. These files will be freely available. Study protocol: Study protocols will not be made available. Statistical code: Computer code used to generate the study results are freely available at https://github.com/rvajravelu/statinDisparities with a README file describing implementation instructions. Data set: All raw data are freely available via the National Center for Health Statistics at www.cdc.gov/nchs/nhanes/index.htm.

RESOURCES