Traditional CHD risk factors were systolic blood pressure (ideal level: ≤110 mm Hg), low-density lipoprotein cholesterol (ideal: ≤70 mg/dL for those with a history of diabetes or cardiovascular disease; ≤100 mg/dL for all others), high-density lipoprotein cholesterol (ideal: ≥50 mg/dL), body mass index (ideal: ≤25), cigarette smoking (ideal: no exposure), and diabetes (ideal: none). The risk associated with low SES was independent of age, sex, and traditional factors.
Each risk factor was simulated in isolation (1) by exchanging the distributions between adults with low socioeconomic status (SES) and their age- and sex-matched counterparts with higher SES and then (2) by improving each factor to its ideal level. Traditional CHD risk factors were systolic blood pressure (SBP; ideal level: ≤110 mm Hg), low-density lipoprotein cholesterol (LDL-C; ideal: ≤70 mg/dL for those with a history of diabetes or cardiovascular disease; ≤100 mg/dL for all others), high-density lipoprotein cholesterol (ideal: ≥50 mg/dL), body mass index (ideal: ≤25), cigarette smoking (ideal: no exposure), and diabetes (ideal: none).
Incident CHD was defined as angina, myocardial infarction, or cardiac arrest as the index event in a population without preexisting cardiovascular disease. Traditional CHD risk factors were systolic blood pressure (ideal level: ≤110 mm Hg), low-density lipoprotein cholesterol (ideal: ≤70 mg/dL for those with a history of diabetes or cardiovascular disease; ≤100 mg/dL for all others), high-density lipoprotein cholesterol (ideal: ≥50 mg/dL), body mass index (ideal: ≤25), cigarette smoking (ideal: no exposure), and diabetes (ideal: none). The risk associated with low SES was independent of age, sex, and traditional factors. The projections assumed that the traditional and low-SES risk factors remained constant over 30 years, along with age- or sex-stratified event and death rates.
eMethods. Overview of CVD Policy Model Structure and Analytic Approach
eTable 1. Comparisons of Selected Cardiovascular Disease Policy Model Simulation Outputs for 2010 (Model Base Year) With National Targets for 2010
eTable 2. Key Input Values for the Low- and Higher-SES CVD Policy Models, With Measures of Variability Indicated for Inputs That Were Varied in Monte Carlo Simulations
eFigure. Schematic of the Cardiovascular Disease Policy Model
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hamad R, Penko J, Kazi DS, et al. Association of Low Socioeconomic Status With Premature Coronary Heart Disease in US Adults. JAMA Cardiol. 2020;5(8):899–908. doi:10.1001/jamacardio.2020.1458
Is low socioeconomic status associated with early myocardial infarction and coronary heart disease mortality, and what is the proportion attributable to traditional risk factors vs other factors?
In this computer simulation study of the US population aged 35 to 64 years, traditional risk factors for coronary heart disease explained 40% of excess events among those with low socioeconomic status, with the remaining 60% attributable to other risk factors. A simulation involving 1.3 million 35-year-olds with low socioeconomic status projected that 250 000 of these people will develop coronary heart disease by age 65 years, which is nearly double the rate projected for individuals with higher socioeconomic status.
These findings suggest that, although addressing traditional risk factors may decrease coronary heart disease, the disparities in disease burden will likely remain unless upstream factors associated with low socioeconomic status are addressed.
Individuals with low socioeconomic status (SES) bear a disproportionate share of the coronary heart disease (CHD) burden, and CHD remains the leading cause of mortality in low-income US counties.
To estimate the excess CHD burden among individuals in the United States with low SES and the proportions attributable to traditional risk factors and to other factors associated with low SES.
Design, Setting, and Participants
This computer simulation study used the Cardiovascular Disease Policy Model, a model of CHD and stroke incidence, prevalence, and mortality among adults in the United States, to project the excess burden of early CHD. The proportion of this excess burden attributable to traditional CHD risk factors (smoking, high blood pressure, high low-density lipoprotein cholesterol, low high-density lipoprotein cholesterol, type 2 diabetes, and high body mass index) compared with the proportion attributable to other risk factors associated with low SES was estimated. Model inputs were derived from nationally representative US data and cohort studies of incident CHD. All US adults aged 35 to 64 years, stratified by SES, were included in the simulations.
Low SES was defined as income below 150% of the federal poverty level or educational level less than a high school diploma.
Main Outcomes and Measures
Premature (before age 65 years) myocardial infarction (MI) rates and CHD deaths.
Approximately 31.2 million US adults aged 35 to 64 years had low SES, of whom approximately 16 million (51.3%) were women. Compared with individuals with higher SES, both men and women in the low-SES group had double the rate of MIs (men: 34.8 [95% uncertainty interval (UI), 31.0-38.8] vs 17.6 [95% UI, 16.0-18.6]; women: 15.1 [95% UI, 13.4-16.9] vs 6.8 [95% UI, 6.3-7.4]) and CHD deaths (men: 14.3 [95% UI, 13.0-15.7] vs 7.6 [95% UI, 7.3-7.9]; women: 5.6 [95% UI, 5.0-6.2] vs 2.5 [95% UI, 2.3-2.6]) per 10 000 person-years. A higher burden of traditional CHD risk factors in adults with low SES explained 40% of these excess events; the remaining 60% of these events were attributable to other factors associated with low SES. Among a simulated cohort of 1.3 million adults with low SES who were 35 years old in 2015, the model projected that 250 000 individuals (19%) will develop CHD by age 65 years, with 119 000 (48%) of these CHD cases occurring in excess of those expected for individuals with higher SES.
Conclusions and Relevance
This study suggested that, for approximately one-quarter of US adults aged 35 to 64 years, low SES was substantially associated with early CHD burden. Although biomedical interventions to modify traditional risk factors may decrease the disease burden, disparities by SES may remain without addressing SES itself.
Coronary heart disease (CHD) is the leading cause of death in the United States.1 Mortality from CHD has been declining for more than 4 decades, in part because of reduced smoking, control of hyperlipidemia and hypertension, and improved treatment of acute CHD events.2 Unfortunately, improvements have been less marked for individuals with low socioeconomic status (SES), who bear a disproportionate share of the CHD burden.2 Coronary heart disease remains the leading cause of mortality in low-income US counties, whereas cancer has the top distinction in high-income counties.3 Numerous studies have demonstrated an inverse association between SES and CHD mortality.4-6 Although some of this excess risk is attributable to a higher burden of traditional risk factors among individuals with low SES, risk factor profiles may not fully account for the observed differences, suggesting that low SES itself and other upstream characteristics may be independent risk factors.7
Understanding what proportion of CHD disparities are attributable to traditional risk factors vs other aspects of SES has important implications for whether future interventions should focus on traditional risk factor management or policies aimed at upstream factors, such as expanding economic or educational opportunities. In this computer simulation study, we quantified the contribution of low SES to the burden of premature (before age 65 years) CHD occurring in US adults aged 35 to 64 years. This quantification provides insight into the magnitude of disparity between adults with low and those with higher SES that would remain even if traditional CHD risk factors were adequately addressed.
This computer simulation used input parameters drawn from several publicly available data sets, secondary analyses of cohort data, and published literature. The individual studies that provided inputs for the present study obtained informed consent from their participants. Ethics approval for the present study was granted by the institutional review board of the University of California, San Francisco.
We used national data to describe the disparity in CHD risk factors and mortality between adults with low SES and those with higher SES. Next, we used the Cardiovascular Disease Policy Model, an established computer simulation of US adults, to estimate excess myocardial infarction (MI) rates and excess CHD deaths experienced by individuals with low SES. We quantified the proportion of these excess events attributable to traditional risk factors compared with the proportion attributable to other upstream factors associated with low SES itself.
The Cardiovascular Disease Policy Model is a computer simulation, state-transition (Markov) model of CHD and stroke incidence, prevalence, mortality, and health care costs in the US population aged 35 years or older (eFigure in the Supplement).8-12 Using inputs from nationally representative databases, longitudinal cohort studies, and natural history studies, the model projected the annual incidence of CHD, stroke, and noncardiovascular death in individuals without cardiovascular disease on the basis of age, sex, and cardiovascular risk factors. Traditional risk factors included exposure to tobacco smoke, systolic blood pressure (SBP), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared), and the presence of type 2 diabetes. For the present study, we also included a variable representing the elevated risk associated with low SES compared with higher SES independent of traditional risk factors. After the initial cardiovascular event, the model characterized the event (angina, myocardial infarction [MI], arrest, or stroke) along with its sequelae in the first 30 days, a period with a heightened probability of repeat events, revascularization procedures, and mortality. The model then projected the number of subsequent MIs, strokes, and cause-specific deaths as a function of age, sex, and cardiovascular history. The risk of new and repeat cardiovascular events was elevated during the year of the incident event. The current version of the model was calibrated to within 1% of events and deaths reported in the 2010 National Hospital Discharge Survey13 and the Vital Statistics of the United States14 (eTable 1 in the Supplement), aligning with the 2010 US Census15 data. Additional details are provided in the eMethods in the Supplement.
We defined low SES as household income below 150% of the federal poverty level or educational level less than a high school diploma, similar to the definitions used in previous studies.4-6,16 The remaining population was described as higher SES. Theoretical work suggests that using multiple measures of SES helps to capture the different individual and societal forces associated with heart disease.17 We identified the size of the populations with low and higher SES using the 2015 American Community Survey.18 We used data from the 2011 to 2016 National Health and Nutrition Examination Survey19 to define SES-stratified distributions of traditional risk factors by age and sex.
The analysis used the model’s risk function, calculated from Framingham data, to model specifications with a Cox proportional hazards regression model for quantifying the association between traditional risk factors and incident CHD, stroke, diabetes, and death from causes other than CHD or stroke.20,21 We used a relative risk of 1.58 (95% CI, 1.31-1.90) for the association of low SES with incident CHD (independent of the traditional risk factors) on the basis of a published analysis of Atherosclerotic Risk in Communities Study data,16 in which the definition of SES matched that in the present study and controlled for all factors included in the model’s risk function. A similar magnitude of association between SES and incident CHD was reported in individuals younger than 65 years in the Reasons for Geographic and Racial Differences in Stroke cohort study.22
We developed separate versions of the model to represent low-SES and higher-SES populations. We restricted the analysis to individuals aged 35 to 64 years because we were interested in describing the role of SES in the development of CHD at younger ages and because SES after age 65 years is complicated by retirement and receipt of Social Security benefits.22 To estimate the current burden of CHD in populations with low SES and higher SES, we conducted simulations from 2015 to 2024, with those who reached age 35 years entering and those who turned age 65 years exiting the simulations each annual cycle. We also estimated the cumulative incidence of CHD in individuals with low SES aged 35 years who were free of cardiovascular disease in 2015, projecting through 2044 when the cohort will reach age 65 years.
We conducted a series of simulations that isolated individual inputs to apportion events in the low-SES population to 1 of the following: (1) the difference in traditional risk factor distributions observed in adults with low SES compared with those with higher SES; (2) the risk observed in individuals with low SES compared with those with higher SES independent of traditional risk factors; (3) the events that could be prevented if risk factors were improved to ideal levels; and (4) the underlying age- and sex-based risk not explained by traditional risk factors or SES. The ideal levels for traditional risk factors were defined as SBP of 110 mm Hg or lower, LDL-C of 70 mg/dL or lower (to convert milligrams per deciliter to millimoles per liter, multiply by 0.0259) for those with previous diabetes or cardiovascular disease and LDL-C of 100 mg/dL or lower for others, HDL-C of 50 mg/dL or higher, BMI of 25 or lower, and no cigarette smoking and diabetes.23,24 Treating to reach ideal levels may not deliver the same outcomes as if these risk factor thresholds were never exceeded in the first place. To compare the potential outcomes of theoretical interventions to eliminate the risk of low SES independent of traditional risk factors with interventions to eliminate traditional risks, we conducted simulations that first exchanged low-SES measures for higher-SES measures of the given risk factor as observed in the National Health and Nutrition Examination Survey. Then, we further improved the risk factor to its ideal level.
We used Monte Carlo simulations to generate 95% uncertainty intervals (UIs) for model estimates. For each simulation, we conducted 1000 iterations, selecting inputs randomly (with replacement) from standard normal distributions scaled to the mean and CI for each varied parameter (eTable 2 in the Supplement). We used the SE estimated from the National Health and Nutrition Examination Survey to vary values for all risk factor inputs for individuals with low SES and those with higher SES. β Coefficients for the association of CHD events with smoking, SBP, LDL-C, HDL-C, and diabetes were varied based on the SE generated from fitted regressions. We assumed log normality for the relative risk of CHD among adults with low SES compared with those with higher SES, deriving the SE from published 95% CIs. Data were analyzed starting in 2015.
Approximately 31.2 million US adults aged 35 to 64 years had low SES in 2015, of whom approximately 16 million (51.3%) were women. Compared with individuals with higher SES, those with low SES experienced 2-fold higher rates of CHD death per 10 000 person-years (men: 8.9 [95% CI, 8.9-9.0] vs 16.6 [95% CI, 16.5-16.8]; women: 3.4 [95% CI, 3.4-3.4] vs 7.5 [95% CI, 7.4-7.6]) (Table 1). Adults with low SES also had a higher burden of traditional risk factors vs those with higher SES, with higher proportions of smoking among men (0.37 vs 0.17) and women (0.29 vs 0.15) with low SES and worse metabolic indicators, particularly among women with low SES (mean [SE] BMI: 30.8 [0.3] vs 29.6 [0.2]; LDL-C: 122.0 [1.6] mg/dL vs 116.8 [1.8] mg/dL; HDL-C, 54.2 [0.6] mg/dL vs 60.4 [1.4] mg/dL; and proportion of diabetes, 0.19 vs 0.10) (Table 1).
Simulations using the Cardiovascular Disease Policy Model accurately reproduced the observed 2-fold increase in rates of early MI per 10 000 person-years (men: 34.8 [95% UI, 31.0-38.8] vs 17.6 [95% UI, 16.0-18.6]; women: 15.1 [95% UI, 13.4-16.9] vs 6.8 [95% UI, 6.3-7.4]) and CHD death per 10 000 person-years (men: 14.3 [95% UI, 13.0-15.7] vs 7.6 [95% UI, 7.3-7.9]; women: 5.6 [95% UI, 5.0-6.2] vs 2.5 [95% UI, 2.3-2.6]) among individuals with low SES compared with sex-matched individuals with higher SES. After accounting for higher levels of traditional risk factors, the simulations estimated that approximately 60% of the 2-fold excess in MIs and CHD deaths in populations with low SES was attributable to the independent risk of SES and other upstream risk factors (Table 2; Figure 1).
Among adults with low SES, 31% of CHD events were attributable to the risk of low SES independent of traditional risk factors (Figure 1). This finding represents the disparity that would remain even if ideal levels of traditional risk factors were achieved. Meanwhile, 53% of all CHD events were attributable to traditional CHD risk factors (Figure 1).
A simulation of theoretical interventions for each individual risk factor revealed that the potential improvement associated with addressing the independent risk of low SES was greater than that for addressing any traditional risk factor alone (Figure 2). Smoking cessation, along with diabetes prevention in women, represented a high-yield target for decreasing the excess burden of CHD in adults with low SES compared with those with higher SES. Reducing BMI in all individuals with low SES to at least 25 would prevent 6.6 MIs and 2.9 CHD deaths per 10 000 person-years, but this decrease would not change the gap in CHD events between individuals with low SES and those with higher SES because of the high rates of obesity in both populations.
Among the simulated closed cohort of 1.3 million adults with low SES aged 35 years in 2015, we estimated that 250 000 (19%) will develop CHD by age 65 years in 2045. Of the cumulative cases, 119 000 (48%) were in excess of what would be expected if the cohort were composed of individuals with higher SES, with 84 000 (70%) of this excess attributable to risk factors associated with low SES independent of traditional risk factors (Figure 3).
One-quarter of the US population aged 35 to 64 years had low SES in 2015 (31.2 million), with rates of MIs and CHD deaths that were twice as high as those among individuals with higher SES. Although a substantial portion of this excess risk can be explained by a higher prevalence of traditional risk factors among adults with low SES, we found that 60% of the excess risk would remain if these factors were brought to the same level as that of the higher-SES population.
Findings of this study suggest that clinical interventions addressing traditional risk factors may decrease the burden of CHD, but they are unlikely to eliminate the socioeconomic disparities in CHD without interventions that directly address other upstream risk factors associated with low SES, such as poverty and education.25 These social conditions have been termed the fundamental causes of disease,26 and at least 3 interrelated pathways have been theorized to explain the association of SES with CHD.27 The first pathway is greater psychosocial stressors. A growing body of literature suggests that the cumulative stress from chronic poverty may be associated with biological changes that are atherogenic, perhaps because of elevated levels of catecholamines and cortisol.28 Allostatic load, a cumulative measure of the body’s response to stress, is also believed to be associated with weathering (premature deterioration of health) and an increased risk of CHD and mortality.29,30 In the United States, in which race/ethnicity and SES are highly connected, discrimination and structural racism may be factors in psychosocial stress, which in turn is associated with increased risk of CHD for black individuals and other non-white groups.31-33
The second pathway is limited economic and educational opportunity, which may reduce access to nutritious food, health care, and safe neighborhoods for physical activity.34-38 Poverty is also believed to play a role in decreased cognitive bandwidth and risky decision-making concerning health behaviors such as smoking, alcohol consumption, and drug use.39-41 The third pathway is social norms; peer influence among individuals with higher SES is associated with healthier lifestyle decisions.42,43 Low SES may also have intergenerational implications: one-third of US children live in poverty, and numerous studies have documented worsened CHD risk factors during childhood among children with low SES,44,45 likely associated with all 3 pathways described.46
For each of these pathways, part of the association of low SES with cardiovascular outcomes may be mediated by traditional risk factors. For instance, individuals with low SES have reduced access to health care, which may be a factor in undertreatment of traditional risk factors. Limited health care access may be linked to the high cost of medical care itself as well as difficulties with less transportation and childcare and poor health literacy.47-49 Nevertheless, as early as the 1950s in the first Framingham study, SES was recognized as an important independent factor in individual risk.50 The Whitehall study of British civil servants, all of whom were employed and had health care access, found an inverse association between occupational status and CHD.51 Even those in the middle of the hierarchy demonstrated increased CHD risk, with stressful work environments and social support implicated as potential mechanisms. Most recently, a mendelian randomization study of more than 500 000 adults demonstrated that traditional risk factors mediated only about one-third of the association between genetic variants for educational attainment and CHD.52 Although few studies have directly examined the possible approaches for addressing the upstream determinants of CHD,53,54 such interventions are supported by this study’s finding that low SES adds to the persistent disparities in CHD beyond its association with traditional risk factors.
Addressing traditional risk factors remains an important target for improving the health of individuals with low SES because these factors account for 53% of overall burden and 40% of excess burden. For example, although tobacco use is declining overall, improvements are less pronounced for adults with low SES, making tobacco control a high-yield target for reducing CHD disparities.55 Targeting traditional risk factors may have a meaningful outcome of narrowing the gap in CHD burden among women given that adverse risk factor profiles explain a larger proportion of excess events in women than in men. Yet socioeconomic factors may be the root of the higher burden of some traditional risk factors; for example, the obesity epidemic disproportionately affects women with low SES,56 and income-related factors, such as food insecurity, are associated with increased metabolic risk.57,58 These observations suggest that efforts to promote smoking cessation, lower LDL-C and SBP, and prevent and treat diabetes may decrease SES disparities, although these efforts should be coupled with interventions targeting upstream social determinants.
Approaches that solely target clinical risk factors may paradoxically increase disparities because groups with higher SES are more likely to reap the rewards of interventions in health care settings.59 Current prevention guidelines from the American College of Cardiology and American Heart Association highlight the importance of screening people for socioeconomic disadvantages that may hamper their ability to afford nutritious food or to engage in physical activity, although no recommendations are given for how to help those with a concerning screen.60,61 High-quality studies are needed to test downstream clinical interventions for minimizing traditional risk factors in low-SES populations (eg, tailored, accessible lifestyle programs or affordable medication regimens) and interventions for improving SES itself (eg, social service referrals facilitated in the clinical encounter).62 The latter could be based on existing experimental and quasiexperimental evidence on programs and policies that address socioeconomic risk factors to improve health outcomes.
Findings of this study also have clinical implications for cardiovascular risk assessment. Current approaches to cardiovascular disease prevention are tiered according to estimated cardiovascular risk.63 Although cardiovascular risk scores in some countries include some measure of SES,64 the risk scores of the American College of Cardiology and American Heart Association do not include SES. However, these guidelines note that the score may underestimate the risk in individuals with low SES.65,66 Race/ethnicity is included in the score, highlighting a challenge of assessing cardiovascular risk in the United States, where race/ethnicity is associated with SES. Without a more complete understanding of the mechanisms that underlie higher risk among black individuals, including race/ethnicity, but not SES in cardiovascular risk assessment may lead to undertreatment of non-black individuals with low SES and overtreatment of black individuals with higher SES. Because of the lack of available inputs in the literature, we were not able to model subgroup associations with SES in different racial/ethnic groups. Future studies could use modeling to estimate this association and disentangle the competing pathways. Similarly, failure to consider neighborhood-level SES may underestimate cardiovascular risk and diminish the success of health interventions.67,68 Internationally, several cardiovascular risk scores include individual- or neighborhood-level socioeconomic deprivation69,70; future studies should develop and test such scores for the United States.
The large size of the low-SES population and the large proportion of excess risk unexplained by traditional risk factors suggest that meeting the public health goals of cardiovascular disease prevention will require a focus on upstream factors associated with SES. Addressing the overall burden of CHD and the factors that explain the disparities in CHD will require research into the 60% excess risk not attributable to traditional risk factors among individuals with low SES and factors that are modifiable, particularly among the broader socioeconomic determinants of health across the life course. Previous experimental and quasiexperimental studies have demonstrated that early childhood educational interventions and policies to increase educational attainment have powerful long-lasting implications for CHD and associated risk factors.53,71,72 Similar experimental and quasiexperimental work has suggested the advantages of socioeconomic interventions to address income and neighborhood deprivation.36,37,73
In addition, policies that directly improve downstream pathways, such as diet and physical activity at the population level, specifically among individuals with low SES, should be evaluated. These interventions may include nutrition assistance programs, such as the Supplemental Nutrition Assistance Program and the Special Supplemental Nutrition Program for Women, Infants, and Children,74,75 or green-space initiatives to promote neighborhood walkability.76
This study has several limitations. The analyses were limited by the uncertainty in model inputs. The association of SES with CHD was estimated from studies of community-based cohorts; although these cohorts were diverse, the generalizability of these findings to the entire US population is uncertain. The association of SES and CHD has been observed in a number of US cohorts, including the Framingham study.50 In contrast, estimates in this study were derived from the Atherosclerosis Risk in Communities study,16 which may have a higher risk of CHD than other cohorts. Previous studies defined SES in different ways, often including combinations of measures of income, education, wealth, and occupation.51,77 Given the data constraints, we included only income and education in this study’s definition of SES, thereby limiting our ability to explore the implications of other aspects of SES. In addition, much of the literature that provided the input for the simulations placed SES into 2 categories, limiting our ability to use a continuous SES exposure to model a more granular dose-response association with CHD. Nevertheless, previous studies have also suggested that the association between SES and CHD is not linear78,79 and that, for education in particular, the attainment of high school and college credentials serves as an important threshold.80 Furthermore, the true implications of risk factors over the lifespan and the variance based on degree of control may not be fully captured. Similarly, the implications of diet and physical activity for CHD, independent of their association with hypertension, lipids, and diabetes, were not captured. We simulated individuals aged 35 to 64 years given our interest in capturing the role of SES in early CHD; the magnitude of these results, however, may not extrapolate to the role of SES in older populations.
We believe that the findings from this study underscore the need to broaden the understanding of CHD risk and to develop interventions that go beyond traditional clinical risk factors. To reduce the socioeconomic disparities in CHD, we must focus future research on the implementation and evaluation of clinical, community-based, and population-level interventions that target upstream risk factors associated with low SES, such as poverty and education.
Accepted for Publication: March 12, 2020.
Corresponding Author: Kirsten Bibbins-Domingo, PhD, MD, MAS, Center for Vulnerable Populations, University of California, San Francisco, 550 16th St, San Francisco, CA 94143 (firstname.lastname@example.org).
Published Online: May 27, 2020. doi:10.1001/jamacardio.2020.1458
Author Contributions: Ms Penko and Dr Coxson had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Hamad, Penko, Mason, Bibbins-Domingo.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Hamad, Penko, Guzman.
Critical revision of the manuscript for important intellectual content: Kazi, Coxson, Wei, Mason, Wang, Goldman, Fiscella, Bibbins-Domingo.
Statistical analysis: Penko, Kazi, Guzman, Wei, Mason.
Obtained funding: Fiscella, Bibbins-Domingo.
Administrative, technical, or material support: Penko, Wei, Goldman.
Supervision: Kazi, Coxson, Wang, Bibbins-Domingo.
Conflict of Interest Disclosures: Dr Goldman reported receiving grants from the Agency for Health Care Policy and Research, Agency for Healthcare Research and Quality, and National Center for Health Services Research; Henry J. Kaiser Family Foundation; National Heart, Lung, and Blood Institute; Flight Attendant Medical Research Institute; Swanson Family Fund; Bristol-Myers Squibb; American College of Cardiology; and National Institutes of Health (NIH) during the conduct of the study as well as personal fees from Elsevier and Little, Brown outside the submitted work. All of the grants were awarded to institutions, not to Dr Goldman, and were used to develop, upgrade, and update the Cardiovascular Disease Policy Model. No other disclosures were reported.
Funding/Support: This work was supported by grants R01HL081066 and K24DK103992 (Dr Bibbins-Domingo) and K08HL132106 (Dr Hamad) from the NIH; Grant-in-Aid 09GRNT2060096 (Dr Bibbins-Domingo) from the American Heart Association Western States Affiliate; and institutional research funds from the University of California, San Francisco.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: