Elevated Cardiac Biomarkers in Relatively Healthy U.S. Adults

Introduction

Cardiovascular disease (CVD) is the leading cause of death in the United States.¹ Prevention of CVD events often relies on early identification of higher risk individuals based on known CVD risk factors. However, subclinical CVD risk may go undetected in individuals without traditional risk factors such as diabetes, hypertension, hyperlipidemia, and obesity.¹ Novel approaches to cardiovascular risk stratification in younger, healthier adults are needed to inform clinical care and CVD prevention.² Whether adults with serologic evidence of subclinical CVD, but without traditional CVD risk factors, are at increased risk of adverse cardiovascular outcomes is not well understood. As a result, there are no current guidelines for managing CVD risk in this population.

We assessed the relationship between subclinical CVD as determined by cardiac biomarkers (high-sensitivity cardiac troponins T and I (hs-cTnT, hs-cTnI) and N-terminal-pro-brain natriuretic peptide (NT-proBNP)) and mortality among healthy adults participating in the National Health and Nutrition Examination Survey (NHANES).

Methods

We included 4,544 adults ages 20 years and older who participated in NHANES 1999-2004. We excluded those with prevalent CVD (coronary artery disease, myocardial infarction, angina, stroke, heart failure), diabetes, hypertension, estimated glomerular filtration rate (eGFR) <60 ml/min/1.73m², urinary albumin-to-creatinine ratio (UACR) ≥30 mg/g, and those who underwent dialysis treatment within the last year. Prevalent CVD was determined by participant self-report. Diabetes was defined by hemoglobin A1c ≥6.4%, self-report, or use of hypoglycemic medications. Hypertension was defined as a blood pressure ≥130/80 mmHg, self-report, or use of antihypertensive medications. In a sensitivity analysis, we further excluded patients with hyperlipidemia (defined by use of a statin or total cholesterol ≥240 mg/dL) and active smoking (N=2,524).

Results

The mean (standard deviation) age was 38.2 (12.8) years, 53.9% of the sample were women and 71.4% were non-Hispanic White (Table 1). The median hs-cTnT was 3.4 ng/L [interquartile range (IQR) 2.6, 4.3] among women and 5.3 ng/L [IQR 4.1, 6.8] among men and the median NT-proBNP was 37.7 pg/mL [IQR 20.0, 68.2]. The median hs-cTnI (Abbott) was 1.26 ng/L [0.81, 1.91], the median hs-cTnI (Ortho) was 0.21 ng/L [0, 0.50], and the median hs-cTnI (Siemens) was 1.18 ng/L [IQR 0.44, 2.52]. Overall crude prevalence of subclinical CVD was 8.7% in the whole population. The prevalence was higher in women (13.0%) compared to men (3.7%). The median follow-up time was 17.6 years and there were 83 CVD mortality events and 441 all-cause mortality events overall. Among those with subclinical CVD, there were 33 CVD mortality events and 138 all-cause mortality events.

In multivariable-adjusted models, each doubling of hs-cTnT was associated with a 49% increased risk of cardiovascular mortality (hazard ratio (HR) 1.49, 95% confidence interval (CI) 1.02-2.17, p=0.04) and a 37% increased risk of all-cause mortality (HR 1.37, 95% CI 1.12-1.68, p=0.003) (Table 2). Both the Abbott and Ortho hs-cTnI assays were significantly associated with both cardiovascular and all-cause mortality. In contrast, hs-cTnI Siemens assay was not significantly associated with either CVD mortality (HR 1.1, 95% CI 0.80-1.63, p=0.47) or all-cause mortality (HR 1.09, 95% CI 0.97-1.22, p=0.16) after adjusting for multiple clinical and demographic variables. Each doubling of NT-proBNP was associated with a 41% increased risk of cardiovascular mortality (HR 1.38, 95% CI 1.09-1.74, p=0.008) and a 23% increased risk of all-cause mortality (HR 1.23, 95% CI 1.12-1.34, p<0.0001).

A sensitivity analysis excluding patients with existing hyperlipidemia and current smoking showed similar findings, with larger effect sizes for the associations between cardiac biomarkers and mortality in some cases (Table S1). Stratifying by subclinical CVD showed attenuation of the relationship between each biomarker and all-cause mortality to non-significance among those without subclinical CVD (Table S2). Subclinical CVD, as a dichotomous variable, was not significantly associated with CVD mortality when defined by hs-cTnT and NT-proBNP or by hs-cTnI alone (Table S3).

Discussion

In this sample of asymptomatic, community-dwelling adults without known CVD, chronic kidney disease, diabetes, or hypertension, nearly 9% had evidence of subclinical CVD. Both hs-cTnT and NT-proBNP were significantly associated with CVD and all-cause mortality. This finding expands prior work showing an association between these biomarkers and mortality among higher risk patients with CVD risk factors (e.g., diabetes, hypertension).^3,5–7

The heterogeneity in the association between high-sensitivity cardiac troponin assays and mortality based on assay used suggests that this biomarker may not be suitable for screening for subclinical CVD in the general population. Indeed, prior literature has suggested that hs-cTnI may represent an entirely different pathophysiologic mechanism than hs-cTnT; where hs-cTnT is associated with cardiac structural changes, hs-cTnI may represent atherosclerotic burden.⁸

Traditional CVD risk factors include diabetes, hypertension, hyperlipidemia, chronic kidney disease, obesity, and smoking and patients are counseled and treated to prevent or control these conditions. However, our findings show even younger patients who maintain good health may still carry occult CVD risk. Monitoring traditional, large macro-level risk factors may be insufficient for risk stratification, and further advances in precision medicine may be needed to identify CVD risk in younger populations.⁹ This finding challenges the conventional approach to identifying and screening higher risk adults, at which point disease may be well-established and more difficult or expensive to treat.² The lack of an association after excluding participants with subclinical CVD suggests that the clinical thresholds for defining subclinical CVD are reasonable. By showing the risk associated with subclinical CVD in a population without traditional CVD risk factors, our findings highlight the need for preventative measures to reduce cardiovascular morbidity and mortality at the population level.¹⁰

This analysis is limited by one time measurement of cardiac biomarkers. Additionally, the ascertainment of all-cause and CVD death in NHANES was based exclusively on ICD codes. A large proportion of patients were actively smoking at the time of their NHANES exam and while the analyses were controlled for smoking status, residual confounding from other lifestyle factors may still be present. Nonetheless, a sensitivity analysis excluding patients with hyperlipidemia and active smoking demonstrated similar results. There are several additional strengths to this manuscript, including the use of a nationally-representative, community dwelling population of asymptomatic adults who would not normally be screened for CVD in clinical settings, and inclusion of covariates that capture socio-demographic determinants of health (food insecurity, health insurance, physical activity).¹¹

In summary, subclinical cardiovascular disease poses higher cardiovascular risk among apparently healthy adults than previously appreciated. Efforts to reduce population-wide CVD should consider novel methods for risk stratification and early intervention to mitigate preventable CVD.

Funding

Research reported in this publication was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under Award Number R38HL143584.

Conflicts of Interest

The authors have no conflicts of interest to disclose.