Unadjusted Troponin Concentrations. Boxplot of troponin concentrations (pg/mL) for RA subjects and controls, demonstrating elevated troponin concentrations in RA patients compared to controls (p = 0.0006). Presented as 25th quartile, median, 75th quartile. Fences drawn to nearest value not exceeding 1.5 interquartile range. Observations beyond fences are plotted. doi:10.1371/journal.pone.0038930.g001 

Context in source publication

Context 1

... with rheumatoid arthritis (RA) experience premature mortality that is largely due to cardiovascular disease [1,2]. Cardiovascular mortality in RA is 50–60% higher compared to the general population, and there is increased prevalence of ischemic heart disease [3–6], There is a markedly higher prevalence of congestive heart failure (CHF) in RA patients. For example, in a longitudinal, population-based study, the 30-year cumulative incidence of CHF in RA was 37.1% compared to 27.7% in controls, and CHF, rather than ischemic heart disease, accounted for the majority of the increase in cardiovascular mortality in RA [7]. We have previously reported increased concentrations of N- terminal pro-brain natriuretic peptide (NT-proBNP) in patients with RA [8]. NT-proBNP is released by myocytes in response to myocardial stretch, and subtle increases may reflect subclinical myocardial dysfunction [9]. Cardiac magnetic resonance imaging has demonstrated decreased left ventricular mass in patients with RA [10]. Thus, RA may predispose to the development of heart failure through a pathophysiology distinct from that of ischemic heart disease; and chronic, indolent myocyte loss may be a central process. Cardiac troponins (cTn) are components of the cardiomyocyte contractile apparatus, and circulating concentrations are elevated in the setting of myocardial injury, such as acute coronary syndromes (ACS) [11]. High-sensitivity (hs) cTn assays allow measurement of troponin concentrations below conventional levels of detection and have revealed a spectrum of circulating cTn concentrations spanning low and high levels in both healthy subjects and in patients with overt cardiovascular disease. Low levels of troponin elevation, below threshold levels used to diagnose ACS, are associated with increased CV mortality, both in patients with diagnosed ischemic heart disease and in the general population [12,13]. There is no information about cardiac troponin concentrations in patients with RA. We hypothesized that concentrations high sensitivity cardiac troponin (hs-cTn), a marker of myocyte injury, may be elevated early in the disease process, before the development of clinical heart failure. The study was approved by the Institutional Review Board (IRB) of Vanderbilt University Medical Center, and all subjects were provided written informed consent. Subjects were participants in an ongoing study of cardiovascular risk in RA, and detailed methods have been described previously [14,15]. One hundred and sixty four eligible patients . 18 years of age who met the American College of Rheumatology classification for RA and 90 control subjects without RA were studied. Individuals with heart failure, defined as current heart failure requiring treatment, or fulfillment of at least 2 of three following criteria; a previous history of heart failure, use of digoxin, or use of a diuretic, were excluded from the study. Patients were recruited from an RA registry, by referral from area rheumatologists, and by local advertisement. Control subjects were recruited from among the patient’s acquaintances, by local advertisement, and from a database of volunteers maintained by the Vanderbilt Clinical Research Center. Clinical and demographic information was acquired via structured interview, questionnaires, physical examination, laboratory tests, electron beam computed tomography (CT), and review of medical records. Current and historic medication use was determined from information provided by patients, and from review of medical records. Blood pressure was recorded as the average of 2 measurements obtained 5 minutes apart after the subject had rested in a supine position for at least 10 minutes. Subjects were considered hypertensive if they were taking antihypertensive agents, or if measured systolic blood pressure was greater than 140 mm Hg and/or diastolic pressure was greater than 90 mm Hg. RA disease activity was assessed using the Disease Activity Score based on evaluation of 28 joints (DAS28). The DAS28 is a validated composite index including a 28 joint count for tenderness and swelling, erythrocyte sedimentation rate (ESR), and the patient’s overall assessment of wellbeing [16]. The Framingham risk score (FRS), a composite score of traditional risk factors that includes blood pressure, smoking status, age, and sex, but not diabetes mellitus, was calculated. Blood was collected after an overnight fast for determination of complete blood count, serum creatinine, total cholesterol, high- density lipoprotein (HDL) cholesterol, triglycerides, homocysteine, and low-density (LDL) cholesterol at the Vanderbilt University Medical Center clinical laboratory. In patients with RA, CRP and Westergren ESR were measured by the hospital laboratory, and medical records reviewed to determine the presence or absence of rheumatoid factor (RF). Before 2003, the laboratory did not use a high-sensitivity CRP assay, and low concentrations were reported as , 3 mg/L. Thus, in 40 patients with CRP reported as , 3 mg/ L, concentrations were re-measured by ELISA (Millipore), and the resulting data were used in the current study analysis. High sensitivity cardiac specific troponin-I (hs-cTn-I) was assayed using the Erenna R System (Singulex Corp, Alameda, CA) by technicians blinded to the clinical data. The limit of detection and the 10% coefficient of variation were 0.1 pg/mL and 0.7 pg/mL, respec- tively. TNFa, interleukin-6 (IL-6), and NT-pro-BNP concentrations were measured by multiplex enzyme-linked immunosorbent assay (Linco Research/Millipore, St. Louis, MO). All subjects underwent tomographic imaging with a c-150 scanner (GE/Imatron, South San Francisco, CA) as described previously [14]. Scans were evaluated by a single investigator blinded to patient clinical status. The degree of calcified plaque was calculated as described previously; the sum of partial scores of all coronary artery lesions provided an overall coronary artery calcium score (CACS) for each patient [13]. Descriptive statistics were calculated as the median with the interquartile range (median [IQR]) or mean 6 SD) according to the distribution of the continuous variables. Wilcoxon rank-sum tests were used to compare continuous variables. Person Chi- square tests were used to compare categorical variables. We assessed the correlation between hs-cTn-I concentration and cardiovascular risk factors (body mass index (BMI), high density lipoprotein (HDL) and low density lipoprotein (LDL) cholesterol, triglycerides (TG), diabetic status, homocysteine, lipoprotein(a), inflammation markers (CRP, IL-6 and TNF a ) with Spearman’s rank correlation test separately for patients with RA and control subjects. To assess the independent association between RA status and hs-cTn-I concentration, we used 4 multivariable linear regression models with a priori defined covariates. Those models adjusted for: Model 1. Age, race, and sex (base model); Model 2. Base model and cardiovascular risk factors (body mass index (BMI), high density lipoprotein (HDL) and low density lipoprotein (LDL) cholesterol, triglycerides (TG), diabetic status, homocysteine, lipoprotein(a), current smoking status, and hypertension (HTN)); Model 3. Base model and markers of inflammation (C reactive protein (CRP), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNFa); and Model 4. Base model, cardiovascular risk factors (as in Model 2), markers of inflammation (as in Model 3), and NT-proBNP. Among the RA group, the association between hs-cTn-I concentrations and CACS was examined using the proportional odds model with CACS as the dependent variable, and were adjusted for age, race, sex, and FRS. Concentrations of hs-cTn-I, triglycerides, homocysteine, CRP, TNF- a , and IL-6 were natural logarithm-transformed to improved normality. All calculations were performed using R version 2.10.0 () and two-sided P values less than 0.05 were considered statistically significant. Because cardiac troponin is renally cleared, the results of the study could be confounded in patients with poor renal function. To account for renal function, the statistical models were analyzed by including the estimated glomerular filtration rate (Modification of Diet in Renal Disease calculation) in the model. Further, to control for the presence of known cardiovascular disease in the study subjects, the statistical models were re-analyzed, excluding patients with a diagnosed cardiovascular disease (22 RA patients, 8 control patients excluded). Patients with RA (n = 164) and control subjects (n = 90) were of similar age and had a similar sex distribution (Table 1). As reported previously in this cohort, NT-BNP, CRP, TNFa, IL-6, homocysteine, and CACS were significantly higher in RA than in control subjects [9,14,15]. A history of ischemic cardiovascular disease (stroke, myocardial infarction, angina) or a coronary procedure (coronary artery bypass graft or angioplasty) was present in 13% of patients with RA and 9% of controls (P = 0.29). High sensitivity cTn-I concentrations were significantly higher in patients with RA (median 1.15 pg/mL [IQR 0.73–1.92] than controls (0.77 pg/mL [0.49–1.28](P , 0.001) (Figure 1). Hs cTn-I remained significantly higher in RA patients after adjustment for age, race, and sex (p = 0.002)(Figure 2- Model 1 ); age, race, sex and cardiovascular risk factors (p = 0.004)(Figure 2- Model 2 ); age, race, sex, and markers of inflammation (p = 0.008)(Figure 2- Model 3 ); and age, race, sex, CV ...