Sex and Race Differences in N-Terminal Pro–B-type Natriuretic Peptide Concentration and Absolute Risk of Heart Failure in the Community

Peder L Myhre; Brian Claggett; Bing Yu; Hicham Skali; Scott D Solomon; Helge Røsjø; Torbjørn Omland; Kerri L Wiggins; Bruce M Psaty; James S Floyd; Elizabeth Selvin; Christie M Ballantyne; Amil M Shah

doi:10.1001/jamacardio.2022.0680

. 2022 Apr 27;7(6):623–631. doi: 10.1001/jamacardio.2022.0680

Sex and Race Differences in N-Terminal Pro–B-type Natriuretic Peptide Concentration and Absolute Risk of Heart Failure in the Community

Peder L Myhre ^1,², Brian Claggett ¹, Bing Yu ³, Hicham Skali ¹, Scott D Solomon ¹, Helge Røsjø ², Torbjørn Omland ², Kerri L Wiggins ⁴, Bruce M Psaty ^4,^5,⁶, James S Floyd ⁴, Elizabeth Selvin ⁷, Christie M Ballantyne ⁸, Amil M Shah ^1,^✉

¹Division of Cardiovascular Medicine, Brigham and Women’s Hospital, Boston, Massachusetts

²Division of Medicine, Akershus University Hospital and University of Oslo, Oslo, Norway

³School of Public Health, The University of Texas Health Science Center at Houston, Houston

⁴Cardiovascular Health Research Unit, Department of Medicine, University of Washington, Seattle

⁵Department of Epidemiology, University of Washington, Seattle

⁶Department of Health Systems and Population Health, University of Washington, Seattle

⁷Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland

⁸Center for Cardiovascular Disease Prevention, Baylor College of Medicine, Houston, Texas

Accepted for Publication: March 8, 2022.

Published Online: April 27, 2022. doi:10.1001/jamacardio.2022.0680

^✉

Corresponding Author: Amil M. Shah, MD, MPH, Division of Cardiovascular Medicine, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02445 (ashah11@partners.org).

Author Contributions: Drs Myhre and Shah 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: Skali, Floyd, Shah.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Myhre, Shah.

Critical revision of the manuscript for important intellectual content: Myhre, Claggett, Yu, Skali, Solomon, Røsjø, Omland, Wiggins, Psaty, Floyd, Selvin, Ballantyne.

Statistical analysis: Myhre, Claggett, Yu, Wiggins, Floyd.

Obtained funding: Røsjø, Psaty, Selvin, Shah.

Administrative, technical, or material support: Solomon.

Supervision: Skali, Solomon, Røsjø, Psaty, Floyd, Shah.

Conflict of Interest Disclosures: Dr Myhre has received grants from the National Institutes of Health/National Heart, Lung, and Blood Institute, and South-Eastern Norway Health Authorities; personal fees from Novartis, Amgen, Novo Nordisk, Boehringer Ingelheim, AstraZeneca, and Bayer; and served on advisory boards for Novartis and Novo Nordisk. Dr Claggett reported personal fees from Amgen, Boehringer Ingelheim, Corvia, Cardurion, MyoKardia, and Novartis outside the submitted work. Dr Skali received consultant fees from Astellas. Dr Solomon received grants from Actelion, Alnylam, Amgen, AstraZeneca, Bellerophon, Bayer, Bristol Myers Squibb, Celladon, Cytokinetics, Eidos, Gilead, GlaxoSmithKline, Ionis, Lilly, Mesoblast, MyoKardia, National Institutes of Health/National Heart, Lung, and Blood Institute, Neurotronik, Novartis, Novo Nordisk, Respicardia, Sanofi Pasteur, and Theracos; and consulting fees from Abbott, Action, Akros, Alnylam, Amgen, Arena, AstraZeneca, Bayer, Boeringer Ingelheim, Bristol Myers Squibb, Cardior, Cardurion, Corvia, Cytokinetics, Daiichi Sankyo, GlaxoSmithKline, Lilly, Merck, MyoKardia, Novartis, Roche Diagnostics, Theracos, Quantum Genomics, Cardurion, Janssen, Cardiac Dimensions, Tenaya, Sanofi Pasteur, Dinaqor, Tremeau, CellProThera, Moderna, American Regent, Sarepta, Lexicon, Anacardio, Akros, and Puretech Health. Dr Røsjø has received speaker honoraria from Novartis and consultancy honoraria from SpinChip Diagnostics and Cardinor AS. Dr Omland has received grants and personal fees from consultancy and/or speaker honoraria from Abbott Diagnostics, Roche Diagnostics, and Novartis, and has received research support from AstraZeneca, Abbott Diagnostics, Novartis, Roche Diagnostics, Thermo Fisher, Singulex, and Biomedica via Akershus University Hospital. Dr Wiggins received grants from the National Institutes of Health. Dr Psaty received grants from the National Institutes of Health and serves on the Steering Committee of the Yale Open Data Access Project funded by Johnson & Johnson. Dr Selvin received grants from the National Institute of Health. Dr Ballantyne is a holder of a provisional patent (61721475) filed by Baylor College of Medicine and Roche, has received grants from the National Institutes of Health, and nonfinancial support from Roche Diagnostics. Dr Shah received research support from Novartis and consulting fees from Philips Ultrasound, grants from the National Institutes of Health, and serves on advisory board and received personal fees from Philips Ultrasound and Johnson & Johnson. No other disclosures were reported.

Funding/Support: The Atherosclerosis Risk in Communities study has been funded in whole or in part with federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, and the Department of Health and Human Services (HHSN268201700001I, HHSN268201700002I, HHSN268201700003I, HHSN268201700005I, HHSN268201700004I). The work for this article was also supported by National Institutes of Health/National Heart, Lung, and Blood Institute grants R01HL135008, R01HL143224, R01HL150342, R01HL14818, and K24HL152008 (Dr Shah), National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases grants K24DK106414, and R01DK089174 (Dr Selvin), and National Institutes of Health/National Heart, Lung, and Blood Institute grant R01HL134320 (Drs Selvin and Ballantyne). Funding support for “Building on GWAS for NHLBI-diseases: the U.S. CHARGE consortium” was provided by the National Health Institutes through the American Recovery and Reinvestment Act of 2009 (ARRA) (5RC2HL102419). The reagents for the N-terminal pro–B-type natriuretic peptide assays were donated by Roche Diagnostics. The Cardiovascular Health Study was supported by contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086, and 75N92021D00006; and grants U01HL080295 and U01HL130114 from the National Heart, Lung, and Blood Institute, with additional contribution from the National Institute of Neurological Disorders and Stroke. Additional support was provided by R01AG023629 from the National Institute on Aging.

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.

^✉

Corresponding author.

PMCID: PMC9047747 PMID: 35476049

This cohort study aims to determine whether physiologic determinants of N-terminal pro–B-type natriuretic peptide concentrations account for sex and race differences and aims to more uniformly predict heart failure risk.

Key Points

Question

What is the absolute risk of incident heart failure or death by N-terminal pro–B-type natriuretic peptide (NT-proBNP) concentration in demographic subgroups defined by sex and race?

Findings

In this subgroup analysis of the ARIC study including 12 750 participants without heart failure at visit 2 and 5191 participants at visit 5, absolute risk of heart failure or death varied substantially by race and sex, such that Black men were at 7-fold higher risk and 3-fold higher risk in midlife and late life, respectively, compared with White women at guideline-recommended NT-proBNP thresholds.

Meaning

Applying NT-proBNP cut points irrespective of sex and race may lead to misclassification of risk in certain demographic subgroups, in particular underestimation of risk in Black men and overestimation of risk in White women.

Abstract

Importance

Sex- and race-based differences in N-terminal pro–B-type natriuretic peptide (NT-proBNP) concentrations are poorly understood. Clinical decisions are often informed by absolute—as opposed to relative—risk, but absolute risk of incident heart failure (HF) associated with NT-proBNP concentration across these important demographic categories is unclear.

Objective

To determine whether physiologic determinants of NT-proBNP concentrations account for sex and race differences, and to more uniformly predict HF risk using NT-proBNP in these demographic subgroups.

Design, Setting, and Participants

In the longitudinal Atherosclerosis Risk in Communities epidemiologic prospective community-based cohort study, the association of NT-proBNP concentration with relative and absolute risk of HF by sex- and race-based categories was assessed at study visit 2 (1990-1992) and study visit 5 (2011-2013) using Cox and Poisson regression. These data were analyzed from June 2018 to October 2021. The contribution of clinical, anthropometric, echocardiographic, and laboratory parameters to sex- and race-based differences in NT-proBNP concentration was assessed at visit 5 using linear regression. Participants included were free of HF in midlife (visit 2; a total of 12 750 participants) and late life (visit 5; a total of 5191 participants).

Exposures

NT-proBNP concentration.

Main Outcomes and Measures

Incident HF or death.

Results

Among the 5191 HF-free participants at visit 5, the mean (SD) age was 76.0 (5.2) years, 2104 (41%) were male, 1043 (20%) were Black, and the median (IQR) NT-proBNP concentration was 124 (64-239) pg/. In both midlife and late life, NT-proBNP concentration was lowest in Black men (median [IQR] concentration: visit 2, 30 [14-67] pg/mL; visit 5, 74 [34-153] pg/mL) and highest in White women (median [IQR] concentration: visit 2, 70 [42-111] pg/mL; visit, 5, 154 [82-268] pg/mL). Sex and race differences in NT-proBNP concentration persisted after accounting for age, income, education, area deprivation index, cardiovascular diseases, left ventricular structure (LV), LV function, LV wall stress, weight and fat mass, and estimated glomerular filtration rate. Substantial differences in the absolute risk of incident HF or death existed across the sex- and race-based categories at any NT-proBNP concentration (eg, 7-fold [rate ratio, 6.7; 95% CI, 4.6-9.9] and 3-fold [rate ratio, 2.7; 95% CI, 1.7-4.1] difference at visit 2 and visit 5, respectively, at guideline-recommended thresholds) with higher risk consistently observed among Black men and lower risk in White women. Results were replicated in a cohort of participants from the Cardiovascular Health Study.

Conclusions and Relevance

In this study, sex- and race-based differences in NT-proBNP persisted after accounting for known physiologic determinants. Absolute risk associated with a given value of NT-proBNP varied substantially by sex and race. Consideration of NT-proBNP values in the context of sex and race allows for more uniform prediction of absolute risk across important demographic subgroups.

Introduction

N-terminal pro–B-type natriuretic peptide (NT-proBNP) is an established biomarker for heart failure (HF) diagnosis and prognosis in both acute and stable settings,^1,2 and higher concentrations of NT-proBNP are associated with excess risk of incident HF and cardiovascular death in general population studies.¹ There is increasing interest in using circulating biomarkers, like NT-proBNP, to predict risk of developing HF, particularly given the emergence of agents that appear particularly efficacious for HF prevention, such as sodium-glucose cotransporter-2 inhibitors.³ Natriuretic peptide (NP)–based screening has a class IIa recommendation for patients at risk of developing HF,² and this strategy has been shown to reduce the incidence of left ventricular (LV) dysfunction and HF.⁴ Furthermore, in the nonacute setting, current guidelines recommend the use of NT-proBNP concentration to aid diagnosing HF,² with a US Food and Drug Administration–approved threshold of less than 125 pg/mL (to convert to nanograms per liter, multiply by 1) among individuals younger than 75 years and of less than 450 pg/mL among individuals 75 years or older, commonly used to rule out chronic HF.⁵ In addition to well-recognized associations of age with higher NT-proBNP concentrations, female sex is associated with up to 165% higher circulating NT-proBNP concentrations compared with male sex, and people who identify as Black demonstrate 30% to 40% lower NT-proBNP concentrations than people who identify as White.^6,7,8,9 Although sex hormones and genetic ancestry may contribute, known determinants of NT-proBNP concentrations, including cardiac structure and function, body composition, kidney function, and burden of cardiovascular comorbidities, also vary by sex and race and ethnicity^{6,7,10,11,12,13,14,15,16} but their contribution to the differences in NT-proBNP by sex and race is largely uncharacterized. Furthermore, while the relative risk of incident HF associated with higher NT-proBNP appears consistent across subgroups defined by sex and race and ethnicity,¹ clinical decision-making often relies on assessments of absolute—as opposed to relative—risk.

We hypothesized that the sex-based and race-based differences in NT-proBNP concentrations are not fully explained by known physiologic determinants, including age, cardiovascular comorbidities, cardiac structure and function, body composition, and kidney function. We further evaluated the extent to which interpreting NT-proBNP concentrations incorporating information on sex and race may more uniformly predict absolute HF risk across these important demographic subgroups in midlife and late life.

Methods

Study Design and Population

The Atherosclerosis Risk in the Communities (ARIC) study is an ongoing, prospective, observational study that initially recruited 15 792 men and women in 4 US communities between 1987 and 1989, as previously described.¹⁷ The study protocol was approved by institutional review boards at each field center, and all participants provided written informed consent. The study followed Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

For analyses in midlife (mean [SD] age, 57.3 [5.7] years), a total of 12 750 participants at study visit 2 (1990-1992) were included after excluding participants who identified as races other than Black or White (n = 42), with prevalent HF (n = 700), or with missing data on NT-proBNP (n = 856; eFigure 1 in the Supplement). For analyses in late life (mean [SD] age, 76.0 [5.2] years), 5191 participants at visit 5 (2011-2013) were included after excluding participants who identified as races other than Black or White (n = 18), with prevalent HF (n = 1112), or with missing data for NT-proBNP (n = 217; eFigure 1 in the Supplement).

Clinical Characteristics

Race was self-reported. At both visit 2 and visit 5, hypertension, diabetes, coronary artery disease (CAD), and atrial fibrillation (AF) were defined based on medical history, participant interview, and medication use, as previously described.¹⁸ Total combined family income, years of education, and area deprivation index were obtained as described in the eMethods in the Supplement. Estimated glomerular filtration rate (eGFR) was calculated from serum creatinine concentration, sex, race, and age using the 2009 Chronic Kidney Disease Epidemiology Collaboration (CKD-epi) equation,¹⁹ and the 2021 race-free CKD-epi equation in a sensitivity analysis.²⁰ At visit 5, bioelectric impedance (Tanita Body Composition Analyzer; TBF-300A) was used to calculate fat mass.²¹ Methods for ascertainment of genetic ancestry are provided in the eMethods in the Supplement.

Echocardiography

Detailed methods of echocardiography at visit 5 have been previously published, including reproducibility metrics.²² Comprehensive echocardiography was performed at all 4 field centers using uniform imaging equipment and a study-specific imaging protocol with quantitative analysis performed at a central echocardiography reading center.²² Meridional wall stress was calculated from cuff systolic blood pressure, LV end-systolic diameter, and end-systolic posterior wall thickness, as previously validated.²³

Blood Sampling and Measurements of NT-proBNP

The variability in NT-proBNP concentrations related to freeze-thaw cycles and frozen storage has previously been described.²⁴ NT-proBNP concentrations were analyzed from serum samples stored at −70 °C using a sandwich elecys proBNP II immunoassay (Roche Diagnostics). For more details on blood sampling and measurement, see the eMethods in the Supplement.

Outcome Measures

Cardiovascular events in ARIC are collected through active surveillance of all cohort participants, as previously described, with events through December 31, 2017, included in this analysis.²⁵ Incident HF was based on first HF hospitalization or HF death according to International Classification of Diseases, Ninth Revision (ICD-9) and ICD-10 codes (code 428.x or I50 in any position).²⁶ All-cause death was obtained from the National Death Index and hospital discharge lists for in-hospital deaths. The primary outcome of interest in this study was incident HF among participants who remained alive, ie, incident HF or all-cause death, during 5 years of follow-up after visit 2 and visit 5.

Replication Cohort

We assessed the generalizability of our findings among 3920 participants in the Cardiovascular Health Study (CHS) who were free of HF and had NT-proBNP measurement performed at the 1992 to 1993 study visit (mean [SD] age, 74.5 [5.2] years; 2382 women [61%], 648 Black participants [17%]). For more details on the replication cohort, see the eMethods in the Supplement.²⁷

Statistical Analysis

We used linear regression to assess whether differences in log-transformed NT-proBNP concentrations by sex and race were accounted for by measures of cardiac structure and function, body composition, kidney function, age, social determinants of health (SDOH [income, education level, area deprivation index]), and comorbidities at visit 5, as described in the eMethods in the Supplement. At both visit 2 and visit 5, the association between log-transformed NT-proBNP and time to incident HF or death was estimated using Cox regression models that were adjusted for demographic characteristics, SDOH, comorbidities (diabetes, hypertension, CAD, and AF), body mass index (BMI), systolic blood pressure, and eGFR. We used Poisson regression to estimate 5-year risk of incident HF or death as a function of NT-proBNP levels for each sex- and race-specific category. Linear and nonlinear models were tested for each category, and the model with the best fit (ie, lowest Bayesian information criterion) was used. We identified NT-proBNP levels (with 95% CIs) associated with 5%, 10%, and 20% absolute risk of incident HF or death at visit 2, and with 10%, 20%, and 30% risk at visit 5 (given the higher overall HF risk later in life) in each category. These NT-proBNP values were then used as the cutoff for each category, and compared with guideline-recommended and US Food and Drug Administration–approved cutoff values for chronic HF (125 pg/mL if younger than 75 years, 450 pg/mL if 75 years or older, irrespective of sex and race) using adjusted Cox regression models. Similar analyses assessing the associations of ARIC-derived NT-proBNP thresholds or guideline-recommended cutoffs with incident HF or death were performed in CHS. A 2-sided P < .05 was considered statistically significant. All statistical analysis was performed between June 2018 and October 2021 using Stata, version 16 (StataCorp).

To comprehensively characterize differences in absolute risk of HF or death across the full age range at visit 2 and visit 5 combined by sex and race, we performed a pooled analysis of the 2 visits with age and NT-proBNP as continuous variables. We used age-adjusted Poisson regression, and tested for linear, cubic, and quadratic terms of age and NT-proBNP to identify the model with the best fit for each sex- and race-specific category and estimated the risk for NT-proBNP concentrations of 50 pg/mL, 125 pg/mL, and 450 pg/mL and ages 50, 60, and 70 years.

Results

Determinants of NT-proBNP and Differences in Concentrations by Sex and Race

Among the 5191 HF-free participants at visit 5, the mean (SD) age was 76.0 (5.2) years, 2104 were male (41%), 1043 were Black (20%), and the median (IQR) NT-proBNP concentration was 124 (64-239) pg/mL. Higher NT-proBNP concentration was associated with older age, White race, female sex, lower socioeconomic status, several cardiovascular comorbidities, lower BMI and eGFR, greater LV mass, worse LV systolic and diastolic function, and higher LV wall stress (Table).

Table. Baseline Characteristics and Echocardiographic Measurements at ARIC Visit 5 by Quartiles of Visit 5 NT-proBNP Concentration Among Participants Free of Heart Failure.

Characteristic	NT-proBNP quartile, mean (SD)				P value for trend^a
Characteristic	First	Second	Third	Fourth	P value for trend^a
Total, No.	1298	1298	1298	1297	NA
NT-proBNP range, pg/mL	<5 to 64.2	64.3 to 124.2	124.3 to 239.0	239.1 to 27643	NA
Age, y	73.7 (4.3)	75.4 (4.8)	76.4 (4.9)	78.5 (5.4)	<.001
Race, No. (%)
Black	413 (31.7)	271 (20.9)	212 (16.3)	152 (11.7)	<.001
White	885 (68.3)	1027 (79.1)	1086 (83.7)	1145 (88.3)
Sex, No. (%)
Male	650 (50.0)	534 (41.1)	431 (33.2)	489 (37.7)	<.001
Female	648 (50.0)	764 (58.9)	867 (66.8)	808 (62.3)
Education level, No. (%)^c
Basic	161 (12.4)	158 (12.2)	143 (11.0)	184 (14.2)	<.001
Intermediate	499 (38.5)	513 (39.6)	601 (46.3)	573 (44.2)
Advanced	637 (49.1)	624 (48.2)	553 (42.6)	538 (41.5)
Family income at visit 1, No. (%)^c
Low	142 (11.6)	154 (12.5)	136 (11.1)	153 (12.3)	.01
Middle	340 (27.7)	356 (29.0)	363 (29.6)	415 (33.4)
High	747 (60.8)	718 (58.5)	727 (59.3)	676 (54.3)
High area deprivation index, No. (%)	251 (19.4)	202 (15.6)	182 (14.1)	128 (9.9)	<.001
History of coronary heart disease, No. (%)	66 (5.2)	105 (8.3)	128 (10.0)	234 (18.4)	<.001
Prevalent atrial fibrillation, No. (%)	20 (1.6)	38 (3.1)	50 (4.1)	209 (17.3)	<.001
Systolic blood pressure, mm Hg	127 (15)	129 (18)	131 (18)	135 (20)	<.001
Hypertension, No. (%)	1019 (78.4)	1024 (78.9)	1041 (80.2)	1146 (88.3)	<.001
Body mass index^b	29.3 (5.2)	28.5 (5.3)	28.0 (5.4)	27.8 (5.7)	<.001
Fat body mass, %	30 (11)	28 (11)	27 (11)	26 (11)	<.001
Diabetes, No. (%)	507 (39.0)	442 (34.1)	412 (31.7)	440 (33.9)	.003
eGFR, mL/min/1.73 m²	76 (15)	73 (15)	70 (16)	63 (18)	<.001
LV end-diastolic volume index, mL/m²	42.1 (9.1)	42.7 (9.7)	42.9 (9.6)	45.4 (11.6)	<.001
Wall thickness, cm	0.98 (0.12)	0.97 (0.12)	0.96 (0.13)	1.01 (0.16)	<.001
LV mass index, g/m^2.7	35.3 (8.2)	36.3 (9.2)	36.7 (9.2)	40.4 (11.4)	<.001
LV ejection fraction, %	65.9 (5.2)	66.4 (5.4)	66.0 (5.6)	64.4 (7.2)	<.001
Global longitudinal strain, %	−18.0 (2.3)	−18.3 (2.2)	−18.3 (2.3)	−17.6 (2.8)	<.001
TDI e', cm/s	5.8 (1.4)	5.8 (1.4)	5.7 (1.5)	5.7 (1.7)	.34
E/e′ ratio	11.3 (3.2)	11.6 (3.4)	12.1 (3.8)	13.6 (5.3)	<.001
LA volume index, mL/m²	22.4 (5.9)	23.7 (6.7)	25.3 (7.1)	30.5 (10.8)	<.001
Meridional wall stress, No. (%)	167 (35)	172 (39)	177 (41)	183 (48)	<.001

Open in a new tab

Abbreviations: ARIC, Atherosclerosis Risk in Communities Study; eGFR, estimated glomerular filtration rate; LA, left atrial; LV, left ventricular; NA, not applicable; NT-proBNP, N-terminal pro–B-type natriuretic peptide; TDI, tissue Doppler imaging.

^{^a}

P values for trend across quartiles of NT-proBNP concentrations.

^{^b}

Calculated as weight in kilograms divided by height in meters squared.

^{^c}

Educational level groups were defined as basic (≤11 years of education), intermediate (12-16 years of education), or advanced (17-21 years of education). Family income categories were defined as low (<$15 999), middle ($16 000-34 999), or high (≥$35 000). See Supplemental Methods for further details.

At visit 5, NT-proBNP was 19% (95% CI, 14%-24%) lower in men compared with women (P < .001). Sex significantly modified the associations between NT-proBNP and history of AF, such that the magnitude of each these associations was stronger in men compared with women (eFigure 2 in the Supplement). After accounting for known determinants of NT-proBNP (age, race, SDOH, comorbidities, blood pressure, BMI, fat mass, diabetes, eGFR, and LV structure and function, including the significant sex-based interactions), NT-proBNP concentrations were 39% (95% CI, 35%-45%) lower in men compared with women (P < .001; Figure 1A).

Figure 1. — Abbreviations: ADI, area deprivation index; BMI, body mass index; eGFR, estimated glomerular filtration rate; GLS, global longitudinal strain; LAVi, left atrial volume index; LV, left ventricle; LVEDVi, left ventricle end-diastolic volume index; LVEf, left ventricle ejection fraction; LVMi, left ventricle mass index; LVWT, left ventricle wall thickness.

Black participants had 38% (95% CI, 33%-42%) lower NT-proBNP concentrations than White participants in unadjusted models at visit 5 (P < .001). Race significantly modified the association of NT-proBNP with LV volume, such that the magnitude of this association was greater in Black participants compared with White participants (eFigure 3 in the Supplement). In multivariable models accounting for known determinants of NT-proBNP, and including an interaction term between race and LV volume, Black race was associated with a 30% (95% CI, 24%-36%) lower NT-proBNP concentration (Figure 1B). The results were similar when using the race-free CKD-epi equation for eGFR (eFigure 4 in the Supplement). There is high consistency between self-identified race and genetic ancestry based on principal component analysis of 17 751 genotyped single-nucleotide variants in ARIC. Sensitivity analysis among 5062 participants with concordant genetic ancestry and self-identified race (993 Black participants, 4069 White participants) demonstrated similar findings with respect to race-based differences in NT-proBNP (eFigure 5 in the Supplement). Additionally, no gradient in race-based differences in NT-proBNP levels were observed when stratifying Black participants by percentage of European ancestry based on 1350 ancestry-informative genetic markers. NT-proBNP levels were 34% (95% CI, 22%-44%), 31% (95% CI, 22%-39%), 25% (95% CI, 12%-35%), and 29% (95% CI, 15%-41%) in Black participants with less than 10%, 10% to 20%, 20% to 30%, and more than 30% European ancestry, respectively, compared with White participants in adjusted models (eFigure 6 in the Supplement).

Sex- and Race-Based Differences in NT-proBNP and HF Risk in Midlife

Among the 12 750 ARIC participants free of HF at visit 2, the mean (SD) age was 57.3 (5.7), years 5670 (45%) were men, and 3129 (25%) were Black . The median (IQR) concentration of NT-proBNP was 30 (14-67) pg/mL for Black men, 46 (23-86) pg/mL for Black women, 39 (22-72) pg/mL for White men, and 70 (42-111) pg/mL for White women (Figure 2A), differences that remained significant after adjusting for demographic characteristics, comorbidities, BMI, systolic blood pressure, and eGFR (eTable 1 in the Supplement).

During 5-year follow-up, 622 participants (4.9%) developed HF or died (incidence rate, 1.02 per 100 person-years [95% CI, 0.94-1.10]; 324 incident HF and 601 died). The incidence rate per 100 person-years was 1.72 (95% CI, 1.41-2.10) for Black men, 1.27 (95% CI, 1.13-1.43) for White men, 1.21 (95% CI, 1.01-1.46) for Black women, and 0.57 (95% CI, 0.48-0.67) for White women. Higher concentrations of NT-proBNP were associated with greater risk of incident HF or death after adjusting for age, sex, race, comorbidities, blood pressure, BMI, and eGFR (hazard ratio [HR] per doubling, 1.60; 95% CI, 1.52-1.68; P < .001), with stronger associations observed among White participants (HR, 1.71; 95% CI, 1.60-1.83) than Black participants (HR, 1.49; 95% CI, 1.38-1.61; P for interaction = .02). There was no effect modification by sex (eTable 2 in the Supplement).

Using sex- and race-specific prediction models based on age and NT-proBNP concentration, the predicted risk associated with a given NT-proBNP concentration varied substantially between sex- and race-based categories. A guideline recommended NT-proBNP threshold was associated with a 7-fold (6.7 [95%CI 4.6-9.9]) difference in predicted rate of incident HF or death for White women compared to Black men. Differences in absolute predicted risk between sex- and race-based categories were greater at higher NT-proBNP concentrations (eTable 3 in the Supplement; Figure 3). The thresholds of NT-proBNP associated with 10% 5-year risk of incident HF or death were 76 pg/mL (95% CI, 53-107) for Black men, 116 pg/mL (95% CI, 95-125) for White men, 188 pg/mL (95% CI, 142-250) for Black women, and 402 pg/mL (95% CI, 312-597) for White women. Compared with a single NT-proBNP cutoff (125 pg/mL), sex- and race-specific NT-proBNP thresholds were associated with more uniform observed incidence of HF or death, but similar model discrimination (0.65 [95 % CI, 0.63-0.67] vs 0.64 [95% CI, 0.62-0.66]; P = .25) and population attributable risk (C statistic, 0.35 [95% CI, 0.30-0.39] vs 0.35 [95% CI, 0.31-0.39]; eTable 3 in the Supplement). Similar results were observed in analyses restricted to participants free of prevalent AF or CAD (n = 8823; eFigure 7A in the Supplement), and when using incident HF alone as an outcome (n = 324 events; eFigure 8A in the Supplement).

Figure 3. — A, The NT-proBNP concentration (95% CI) associated with 5-year risk of heart failure or death at 5% was 24 (15-35) and NT-proBNP was 14 pg/mL among Black men; 72 (43-101) and 69 pg/mL among Black women; 51 (41-58) and 49 pg/mL among White men; and 205 (170-255) and 200 pg/mL among White women, respectively. The NT-proBNP concentration (95% CI) associated with 5-year risk of heart failure or death at 10% was 76 (53-107) and NT-proBNP was 70 pg/mL among Black men; 188 (142-250) and was 177 pg/mL among Black women; 116 (95-125) and 109 pg/mL among White men; and 402 (312-597) and 388 pg/mL among White women, respectively. The NT-proBNP concentration (95% CI) associated with 5-year risk of heart failure or death at 20% was 252 (177-386) and NT-proBNP was 213 pg/mL among Black men; 488 (353-758) and 419 pg/mL among Black women; 273 (209-281) and 241 pg/mL among White men; and 813 (557-1744) and 724 pg/mL among White women, respectively. B, The NT-proBNP concentration (95% CI) associated with 5-year risk of heart failure or death at 10% was 31 (14-56) and NT-proBNP was 31 pg/mL among Black men; 86 (55-121) and NT-proBNP was 80 pg/mL among Black women; 87 (67-108) and NT-proBNP was 81 pg/mL among White men; and 141 (112-171) and NT-proBNP was 131 pg/mL among White women, respectively. The NT-proBNP concentration (95% CI) associated with 5-year risk of heart failure or death at 20% was 103 (62-158) and NT-proBNP was 89 pg/mL among Black men; 245 (176-365) and NT-proBNP was 208 pg/mL among Black women; 238 (198-284) and NT-proBNP was 205 pg/mL among White men; and 410 (339-516) and NT-proBNP was 350 pg/mL among White women. The NT-proBNP concentration (95% CI) associated with 5-year risk of heart failure or death at 30% was 209 (137-365) and NT-proBNP was 165 pg/mL among Black men, 451 (317-863) and NT-proBNP was 369 pg/mL among Black women, 445 (367-558) and NT-proBNP was 353 pg/mL among White men, and 798 (609-1167) and NT-proBNP was 625 pg/mL among White women.

Sex- and Race-Based Differences in NT-proBNP and Predicted Risk in Late Life

At visit 5 (mean [SD; range] age, 76.0 [5.2; 67-91] years), the median (IQR) NT-proBNP concentration among the 5188 participants free of HF was 74 (34-153) pg/mL for Black men, 88 (47-165) pg/mL for Black women, 111 (59-237) pg/mL for White men, and 154 (82-268) pg/mL for White women (Figure 2B). These differences remained significant after adjusting for possible confounders (eTable 1 in the Supplement).

At 5-year follow-up, 709 participants (13.7%) developed HF or died (incident rate per 100 person-years, 3.07; 95% CI, 2.84-3.30; 308 with incident HF and 529 died). The incidence rate per 100 person-years was 4.54 (95% CI, 3.58-5.76) for Black men, 3.50 (95% CI, 3.12-3.95) for White men, 2.67 (95% CI, 2.15-3.31) for Black women, and 2.65 (95% CI, 2.36-2.98) for White women. Higher NT-proBNP concentrations were associated with greater risk of incident HF or death in adjusted models (HR per doubling, 1.55; 95% CI, 1.46-1.65; P < .001), with no evidence of effect modification by sex or race (eTable 2 in the Supplement). The absolute predicted risk at any given NT-proBNP concentration varied substantially between sex- and race-based categories. A guideline-recommended NT-proBNP threshold was associated with a 3-fold lower difference in predicted rate of incident HF or death for White women compared with Black men (incidence rate per 100 person-years, 3.7 [95% CI, 3.1-4.5] vs 9.9 [95% CI, 6.8-14.6]; eTable 3 in the Supplement). This difference became greater at higher NT-proBNP concentrations (Figure 3). The NT-proBNP thresholds associated with 20% risk of HF at 5 years were 103 pg/mL (95% CI, 62-158) for Black men, 238 pg/mL (95% CI, 198-284) for White men, 245 pg/mL (95% CI, 176-365) for Black women, and 410 pg/mL (95% CI, 339-516) for White women. Using sex- and race-specific NT-proBNP thresholds was associated with more uniform observed incidence rates across groups, higher risk of incident HF, better model discrimination (C statistic, 0.63 [95% CI, 0.61-0.65] vs 0.59 [95% CI, 0.57-0.61]; P < .001), and greater population attributable risk (0.34 [95% CI, 0.29-0.38] vs 0.24 [95% CI, 0.29-0.38]; eTable 3 in the Supplement). Differences by sex and race were similar in analyses restricted to participants free of AF or CAD (n = 4403; eFigure 7B in the Supplement) and when using incident HF alone as outcome (n = 308 events; eFigure 8B in the Supplement).

Among 3920 CHS participants free of HF, NT-proBNP levels above guideline-recommended thresholds were associated with a 4-fold lower difference in the incidence rate of HF or death for White women compared with Black men (incidence rate per 100 person-years, 2.8 [95% CI, 2.4-3.3] vs 11.0 [95% CI, 7.8-15.6]; rate ratio, 4.0; 95% CI, 2.7-5.8; eTable 4 in the Supplement), with intermediate incidence rates among Black women and White men. Use of the sex- and race-specific NT-proBNP thresholds identified in ARIC substantially attenuated the between-group differences in incidence of HF or death (rate ratio for Black men vs White women, 1.8; 95% CI, 1.3-2.5).

Absolute Risk of HF or Death Associated With NT-proBNP Concentrations Across Midlife to Late Life

Given the recognized association of older age with higher NT-proBNP concentrations, we more fully characterize the differences in absolute risk of HF or death associated with NT-proBNP concentrations at 3 specific ages in midlife to late life in subgroups defined by sex and race (Figure 4). We observed a similar pattern of absolute risk by sex and race subgroups at each age and NT-proBNP concentration.

Discussion

Substantial differences in NT-proBNP concentrations were observed by sex and race in both midlife and late life and were not explained by known determinants of NT-proBNP concentrations, including age, SDOH, cardiovascular comorbidities, LV function or wall stress, body mass and composition, and kidney function. Sex- and race-based differences in NT-proBNP levels resulted in large differences in absolute risk of HF at any given NT-proBNP concentration, absolute differences that become larger at higher NT-proBNP concentrations. Applying a single NT-proBNP cut point irrespective of sex and race therefore may lead to misclassification of risk in certain demographic subgroups, in particular underestimation of risk in Black men and overestimation of risk in White women. Important sex- and race-based differences in HF risk exist, and measurement of NT-proBNP is not a reliable equalizer of risk across these important demographic subgroups. Consideration of sex and race in interpreting NT-proBNP values allows for more uniform prediction of absolute risk across sexes and races.

Differences in circulating NP concentrations by sex and race and ethnicity are well recognized, and contributions from sex hormones and genetic ancestry have been proposed.^6,7 Pro–B-type NP are produced primarily in response to elevated LV wall stress. However, the cardiac structural and functional determinants of LV wall stress differ by sex such that women demonstrate lower LV mass and volumes, higher LVEF,²⁸ greater hypertrophy, and preservation of systolic function in response pressure load.^10,11 In the setting of hypertension, greater diastolic dysfunction and LV hypertrophy have been observed in Black compared with White individuals in the US.^29,30 Importantly, however, race-based differences in NT-proBNP levels persisted, and sex-based differences were accentuated, in our analysis after accounting for LV mass, wall stress, and systolic and diastolic function. The differences also persisted after accounting for income, education level, and area deprivation index, although other social drivers of health may contribute to the differences in NT-proBNP. Additionally, differences in NT-proBNP levels between Black and White participants were generally consistent across categories of genetically determined European ancestry. These findings do not support genetic ancestry as explanatory of the observed race-based differences in NT-proBNP concentrations.

Despite the largely consistent relative risk of incident HF associated with NT-proBNP concentrations across sex and race categories, the absolute risk associated with a given NT-proBNP concentration varied substantially by race and sex in both ARIC and CHS. These absolute differences became larger at higher NT-proBNP concentrations. As attention increasingly focuses on HF prevention,³ there is a growing need to identify persons at elevated risk of HF to target for preventive therapies. While multicomponent risk prediction models have been developed, NPs are also promising tools that could be easily used in clinical practice. In this setting, thresholds based on absolute predicted risk could potentially be used for instituting preventive therapies, similar to current guidelines for primary prevention of CAD. Importantly, however, our findings suggest that assuming the same association between NT-proBNP and absolute risk—or a universal NT-proBNP cutoff —irrespective of sex and race may lead to overestimation of risk for some (in particular White women) and underestimation in others (in particular Black men). We found large variations in sex- and race-tailored NT-proBNP cutoffs to define a specific 5-year risk of HF or death, with median values for White women many-fold higher than for Black men.

Known and unknown variants in genes and biologic pathways influencing NT-proBNP, which we were not able to assess, may contribute to differences by sex and race categories. More importantly, detailed measures of the SDOH that differentially condition lived experience—and possibly activation of specific biologic pathways—by sex and race likely also contribute to these differences. Future studies incorporating detailed measures of the range of social drivers of health to explain interindividual differences in NP concentrations offer promise of a precision medicine approach to using NPs for risk prediction. Until such data are available, interpreting NT-proBNP values in the context of sex and race may improve the uniformity of predicted HF risk across important demographic subgroups in the population.

Limitations

This study has several limitations. First, 1290 of the original ARIC participants (8%) at visit 2 and 3741 participants (36%) at visit 5 were alive but did not attend the visit. However, the clinical characteristics at the first ARIC visit were comparable among living participants who attended compared with those who did not attend (eTables 5-6 in the Supplement). Second, the number of Black participants was limited, which may have limited our statistical power in comparing outcomes in this group with other demographic groups. Third, our analysis studied participants in a community-based prospective cohort and is not necessarily generalizable to symptomatic persons presenting for medical attention. Fourth, we do not have measurements of biologically active NPs, such as B-type NP and A-type NP. Additionally, as NT-proBNP concentrations demonstrate a continuous association with subsequent risk, the use of cut points in this context may be viewed as outdated. Importantly, we did not aim to evaluate existing cut points or to propose new ones, but instead to illustrate the variability in the association between NT-proBNP concentration and absolute risk by sex and race.

Conclusions

In this cohort study, NT-proBNP concentrations varied substantially by sex and race, and these differences were not fully explained by known physiologic determinates of NT-proBNP. Findings were similar when evaluating groups based on self-reported Black race or on African genetic ancestry, which largely overlapped in the study population. The risk of HF or death associated with the guideline-recommended threshold for NT-proBNP was 3-fold to-7-fold higher in Black men compared with White women. Important sex- and race-based differences in risk of HF or death exist and NT-proBNP is not a reliable equalizer of risk across these important demographic subgroups. Consideration of NT-proBNP values in the context of sex and race allows for more uniform prediction of absolute risk across sexes and races.

Supplement.

eMethods.

eReferences.

eTable 1. NT-proBNP levels in black versus white, and women versus men, free of heart failure

eTable 2. Risk (hazard ratio, 95% confidence interval) of incident HF or death per doubling of NT-proBNP adjusted for age, sex/race, body mass index, coronary artery disease, atrial fibrillation, hypertension, diabetes mellitus, systolic blood pressure and estimated glomerular filtration rate.

eTable 3 Comparison of guideline-based NT-proBNP cut-offs versus ARIC-based sex- and race-specific cut-offs in defining risk for incident heart failure or death among participant sub-groups defined by sex and race by Poisson regression.

eTable 4. Comparison of guideline-based NT-proBNP cut-offs versus ARIC-based sex- and race-specific cut-offs in defining risk for heart failure or death among participant sub-groups defined by sex and race using Poisson regression in the Cardiovascular Health Study.

eTable 5. Baseline characteristics at ARIC Visit 1 of participants who were alive at ARIC visit 2, stratified by attendance at ARIC Visit 2.

eTable 6. Baseline characteristics at ARIC Visit 1 of participants who were alive at ARIC visit 5, stratified by attendance at ARIC Visit 5.

eFigure 1. Flow chart of ARIC participants included in the analysis from Visit 2 and Visit 5.

eFigure 2. Association between left ventricular end diastolic volume index and log transformed NT-proBNP concentrations for Black and White participants.

eFigure 3. Association between A) left ventricular mass index and log transformed NT-proBNP concentrations and B) E/e’-ratio and log transformed NT-proBNP for men and women.

eFigure 4. Difference in NT-proBNP between Black and White participants using multivariable linear regression models with additive adjustment using the race free CKD-epi eGFR equation.

eFigure 5. Difference in NT-proBNP between Black and White participants with genetically determined ancestry data available (98%, i.e. n=993 Black participants and n=4,069 White participants) using multivariable linear regression models with additive adjustment.

eFigure 6. Difference in NT-proBNP between Black and White participants, when stratified by percent European ancestry among blacks using multivariable linear regression models with additive adjustment.

eFigure 7. Predicted risk of incident HF or death post-Visit 2 (Panel A) and post-Visit 5 (Panel B) by NT-proBNP concentrations among sex- and race-based categories in participants free of atrial fibrillation and coronary heart disease.

eFigure 8. Predicted risk of incident HF post-Visit 2 (Panel A) and post-Visit 5 (Panel B) by NT-proBNP concentrations among sex- and race-based categories.

Click here for additional data file.^{(1.9MB, pdf)}

References

1.Wang TJ, Larson MG, Levy D, et al. Plasma natriuretic peptide levels and the risk of cardiovascular events and death. N Engl J Med. 2004;350(7):655-663. doi: 10.1056/NEJMoa031994 [DOI] [PubMed] [Google Scholar]
2.Yancy CW, Jessup M, Bozkurt B, et al. 2017 ACC/AHA/HFSA focused update of the 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Task Force on clinical practice guidelines and the Heart Failure Society of America. Circulation. 2017;136(6):e137-e161. doi: 10.1161/CIR.0000000000000509 [DOI] [PubMed] [Google Scholar]
3.Pandey A, Vaduganathan M, Patel KV, et al. Biomarker-based risk prediction of incident heart failure in pre-diabetes and diabetes. JACC Heart Fail. 2021;9(3):215-223. doi: 10.1016/j.jchf.2020.10.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Ledwidge M, Gallagher J, Conlon C, et al. Natriuretic peptide-based screening and collaborative care for heart failure: the STOP-HF randomized trial. JAMA. 2013;310(1):66-74. doi: 10.1001/jama.2013.7588 [DOI] [PubMed] [Google Scholar]
5.Food and Drug Administration. 510(k) Substantial equivalence determination decision summary assay only template. Accessed March 18, 2022. https://www.accessdata.fda.gov/cdrh_docs/reviews/K063662.pdf
6.Lam CSP, Cheng S, Choong K, et al. Influence of sex and hormone status on circulating natriuretic peptides. J Am Coll Cardiol. 2011;58(6):618-626. doi: 10.1016/j.jacc.2011.03.042 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Gupta DK, Claggett B, Wells Q, et al. Racial differences in circulating natriuretic peptide levels: the Atherosclerosis Risk in Communities Study. J Am Heart Assoc. 2015;4(5):e001831. doi: 10.1161/JAHA.115.001831 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Hildebrandt P, Collinson PO, Doughty RN, et al. Age-dependent values of N-terminal pro-B-type natriuretic peptide are superior to a single cut-point for ruling out suspected systolic dysfunction in primary care. Eur Heart J. 2010;31(15):1881-1889. doi: 10.1093/eurheartj/ehq163 [DOI] [PubMed] [Google Scholar]
9.Ballo P, Betti I, Barchielli A, et al. Prognostic role of N-terminal pro-brain natriuretic peptide in asymptomatic hypertensive and diabetic patients in primary care: impact of age and gender: results from the PROBE-HF study. Clin Res Cardiol. 2016;105(5):421-431. doi: 10.1007/s00392-015-0937-x [DOI] [PubMed] [Google Scholar]
10.Hayward CS, Webb CM, Collins P. Effect of sex hormones on cardiac mass. Lancet. 2001;357(9265):1354-1356. doi: 10.1016/S0140-6736(00)04523-2 [DOI] [PubMed] [Google Scholar]
11.Krumholz HM, Larson M, Levy D. Sex differences in cardiac adaptation to isolated systolic hypertension. Am J Cardiol. 1993;72(3):310-313. doi: 10.1016/0002-9149(93)90678-6 [DOI] [PubMed] [Google Scholar]
12.Newton-Cheh C, Larson MG, Vasan RS, et al. Association of common variants in NPPA and NPPB with circulating natriuretic peptides and blood pressure. Nat Genet. 2009;41(3):348-353. doi: 10.1038/ng.328 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Gupta DK, de Lemos JA, Ayers CR, Berry JD, Wang TJ. Racial differences in natriuretic peptide levels: the Dallas Heart Study. JACC Heart Fail. 2015;3(7):513-519. doi: 10.1016/j.jchf.2015.02.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Bajaj NS, Gutiérrez OM, Arora G, et al. Racial differences in plasma levels of N-terminal pro-B-type natriuretic peptide and outcomes: the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study. JAMA Cardiol. 2018;3(1):11-17. doi: 10.1001/jamacardio.2017.4207 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Seidelmann SB, Vardeny O, Claggett B, et al. An NPPB promoter polymorphism associated with elevated N-terminal pro-B-type natriuretic peptide and lower blood pressure, hypertension, and mortality. J Am Heart Assoc. 2017;6(4):e005257. doi: 10.1161/JAHA.116.005257 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Whitman IR, Vittinghoff E, DeFilippi CR, et al. NT-pro BNP as a mediator of the racial difference in incident atrial fibrillation and heart failure. J Am Heart Assoc. 2019;8(7):e010868. doi: 10.1161/JAHA.118.010868 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.The Atherosclerosis Risk in Communities (ARIC) study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129(4):687-702. doi: 10.1093/oxfordjournals.aje.a115184 [DOI] [PubMed] [Google Scholar]
18.Folsom AR, Yamagishi K, Hozawa A, Chambless LE; Atherosclerosis Risk in Communities Study Investigators . Absolute and attributable risks of heart failure incidence in relation to optimal risk factors. Circ Heart Fail. 2009;2(1):11-17. doi: 10.1161/CIRCHEARTFAILURE.108.794933 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Levey AS, Stevens LA, Schmid CH, et al. ; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) . A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. doi: 10.7326/0003-4819-150-9-200905050-00006 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Inker LA, Eneanya ND, Coresh J, et al. ; Chronic Kidney Disease Epidemiology Collaboration . New creatinine and cystatin C-based equations to estimate GFR without race. N Engl J Med. 2021;385(19):1737-1749. doi: 10.1056/NEJMoa2102953 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Valente MA, Hillege HL, Navis G, et al. The Chronic Kidney Disease Epidemiology Collaboration equation outperforms the Modification of Diet in Renal Disease equation for estimating glomerular filtration rate in chronic systolic heart failure. Eur J Heart Fail. 2014;16(1):86-94. doi: 10.1093/eurjhf/hft128 [DOI] [PubMed] [Google Scholar]
22.Shah AM, Cheng S, Skali H, et al. Rationale and design of a multicenter echocardiographic study to assess the relationship between cardiac structure and function and heart failure risk in a biracial cohort of community-dwelling elderly persons: the Atherosclerosis Risk in Communities study. Circ Cardiovasc Imaging. 2014;7(1):173-181. doi: 10.1161/CIRCIMAGING.113.000736 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.DePace NL, Ren JF, Iskandrian AS, Kotler MN, Hakki AH, Segal BL. Correlation of echocardiographic wall stress and left ventricular pressure and function in aortic stenosis. Circulation. 1983;67(4):854-859. doi: 10.1161/01.CIR.67.4.854 [DOI] [PubMed] [Google Scholar]
24.Nowatzke WL, Cole TG. Stability of N-terminal pro-brain natriuretic peptide after storage frozen for one year and after multiple freeze-thaw cycles. Clin Chem. 2003;49(9):1560-1562. doi: 10.1373/49.9.1560 [DOI] [PubMed] [Google Scholar]
25.Loehr LR, Rosamond WD, Chang PP, Folsom AR, Chambless LE. Heart failure incidence and survival (from the Atherosclerosis Risk in Communities study). Am J Cardiol. 2008;101(7):1016-1022. doi: 10.1016/j.amjcard.2007.11.061 [DOI] [PubMed] [Google Scholar]
26.Rosamond WD, Chang PP, Baggett C, et al. Classification of heart failure in the Atherosclerosis Risk in Communities (ARIC) study: a comparison of diagnostic criteria. Circ Heart Fail. 2012;5(2):152-159. doi: 10.1161/CIRCHEARTFAILURE.111.963199 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Fried LP, Borhani NO, Enright P, et al. The Cardiovascular Health Study: design and rationale. Ann Epidemiol. 1991;1(3):263-276. doi: 10.1016/1047-2797(91)90005-W [DOI] [PubMed] [Google Scholar]
28.Gardin JM, Siscovick D, Anton-Culver H, et al. Sex, age, and disease affect echocardiographic left ventricular mass and systolic function in the free-living elderly. the Cardiovascular Health Study. Circulation. 1995;91(6):1739-1748. doi: 10.1161/01.CIR.91.6.1739 [DOI] [PubMed] [Google Scholar]
29.Sharp A, Tapp R, Francis DP, et al. Ethnicity and left ventricular diastolic function in hypertension an ASCOT (Anglo-Scandinavian Cardiac Outcomes Trial) substudy. J Am Coll Cardiol. 2008;52(12):1015-1021. doi: 10.1016/j.jacc.2008.04.065 [DOI] [PubMed] [Google Scholar]
30.Kizer JR, Arnett DK, Bella JN, et al. Differences in left ventricular structure between Black and White hypertensive adults: the Hypertension Genetic Epidemiology Network study. Hypertension. 2004;43(6):1182-1188. doi: 10.1161/01.HYP.0000128738.94190.9f [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials