Factors Related to Silent Myocardial Damage in Hemodialysis Patients

Abstract

Background. Both traditional and non-traditional risk factors play a role for the development of cardiovascular disease in hemodialysis patients. However, a specific relationship between these risk factors and silent myocardial damage is unknown. Methods. Demographic, anthropometric, clinical, and laboratory data were collected. Silent myocardial damage was defined by elevated cardiac troponin I values above cutoff values. Results. In total, 113 hemodialysis patients were included. Cardiac troponin I concentrations were below cutoff value (<2.3 ng/mL) in 103 (91.2%) patients (Group 1), whereas 10 (8.8%) patients had elevated concentrations (Group 2). Group 1 patients had higher levels of hemoglobin (p = 0.002) and high-density lipoprotein cholesterol (p = 0.002) and lower C-reactive protein (p = 0.003) and tumor necrosis factor-α (p = 0.005) levels, as well as less incidence of left ventricular hypertrophy (p = 0.045), when compared to Group 2 patients. Diabetes mellitus (Beta = +0.160, p = 0.021), left ventricular hypertrophy (Beta = +0.247, p < 0.0001), uncontrolled blood pressure (Beta = +0.170, p = 0.016), normalized protein equivalent of total nitrogen appearance (Beta = −0.230, p = 0.001), hemoglobin (Beta = −0.302, p < 0.0001), and tumor necrosis factor-α (Beta = +0.506, p < 0.0001) were found to be independently associated with cardiac troponin I levels in multiple linear regression analysis. Conclusions. Both traditional and non-traditional risk factors are related with silent myocardial damage, which is considered to an antecedent of major cardiovascular events. Hemodialysis patients, even when asymptomatic, must be closely followed up for the presence of these risk factors.

Keywords:

• myocardial damage
• hemodialysis
• traditional
• non-traditional
• troponin I
• tumor necrosis factor alpha

INTRODUCTION

The incidence of cardiovascular death is 5–10 times greater in uremic patients than in the age-matched general population, and cardiac death is related to both traditional risk factors (e.g., advanced age, smoking, diabetes mellitus, hypertension, and dyslipidemia) as well as non-traditional ones (e.g., inflammation, malnutrition, oxidative stress, vascular calcification, and homocysteine).^[1,2] Diagnosing cardiac disease in end-stage renal disease (ESRD) patients is difficult, because many of these patients present with abnormal baseline electrocardiograms, frequently compounded by silent or atypical symptoms, and reduced reliability of cardiac enzymes, including cardiac troponin T (cTnT).^[3] Cardiac troponin I (cTnI) is exclusively of cardiac origin, and unlike cTnT, is not expressed in skeletal muscle. It is suggested that cTnI is a sensitive marker of myocardial injury and is more specific for the detection of myocardial injury in ESRD patients than other cardiac enzymes.^[3–5] Additionally, among the troponins, especially cTnI values are found to be clearly associated with silent/subtle myocardial damage.^[6] Although there is considerable controversy regarding the significance of elevated cTnI levels in asymptomatic hemodialysis patients, there are studies showing that increased cTnI is a risk factor for future cardiac events, including myocardial infarction, cardiac death and revascularization.^[7–10] Interestingly, higher albeit normal troponin concentrations are considered to be markers of active but asymptomatic coronary artery disease, which is likely to progress to an acute coronary syndrome in relatively near future.^[10] Thus, a high level of troponin may not only be observed due to clinically manifest cardiovascular disease, but it might also be seen in silent/subtle myocardial damage caused by microvascular lesions or direct injury to myocardial cells (e.g., toxic, stretching, hypoxia, apoptosis). Indeed, silent/subtle myocardial damage might be considered as a precursor of more severe and clinically detectable cardiac events.^[11]

In this study, we analyzed the potential risk factors (both traditional and non-traditional) for silent/subtle myocardial damage, as defined by elevated cTnI values, in asymptomatic ESRD patients on regular hemodialysis.

MATERIAL AND METHODS

This cross-sectional study was undertaken in the hemodialysis unit of Baskent University Hospital. The study was in accordance with the declaration of Helsinki, and written informed consent was obtained from all patients before enrollment. None of the patients had chest pain, electrocardiographic changes suggestive of arrhythmia, myocardial injury, or ischemia, except for left ventricular hypertrophy. None of the patients experienced intradialytic hypotension, cramps or a bleeding episode during the hemodialysis sessions within last four weeks. Patients with acute infections, active hepatitis, or non-specific fever, as well as those receiving nutritional support, were not included. Any patient suffering from acute coronary syndrome, myocardial infarction, angina pectoris, or coronary revascularization procedure (coronary stent replacement and coronary artery bypass graft surgery) and patients with a history of stroke, transient ischemic attack, carotid revascularization procedure, intermittent claudication, ischemic leg ulcer, peripheral vascularization, or amputation for critical limb ischemia within the last three months were excluded.

Baseline demographic, anthropometric, and clinical data including smoking history, body mass index, presence of diabetes mellitus, and use of antihypertensive medication and 3-hydroxy-3-methyl-glutaryl-CoA reductase inhibitors (statins) were recorded. Patients were considered as having atherosclerosis if at least one of the following clinical conditions were present: coronary artery disease (defined as previous myocardial infarction, angina pectoris, or coronary revascularization procedure), peripheral vascular disease (defined as previous intermittent claudication, ischemic leg ulcer, peripheral vascularization, or amputation for critical limb ischemia), cerebrovascular disease (defined as previous stroke, transient ischemic attack, or carotid revascularization procedure). All patients were evaluated by physical examination, and pre-dialysis blood pressure measurements were obtained. Dry weight was established for each patient on a trial-and-error basis and was defined as the weight below which the patient suffered frequent hypotensive episodes during the latter part of the dialysis session and experienced malaise, cramps, and dizziness postdialysis.

Uncontrolled hypertension was defined as pre-dialysis systolic blood pressure >140 mmHg and/or pre-dialysis diastolic blood pressure >90 mmHg. Electrocardiography was obtained before the hemodialysis session. Left ventricular hypertrophy was electrocardiographically determined by using Sokolow-Lyon voltage (sum of the amplitude of the S wave in lead V₁ and the R wave in lead V₅ or V₆ >35 mm).^[12]

All patients were receiving four-hour hemodialysis thrice weekly with standard bicarbonate dialysis solution. The adequacy of dialysis was assessed by single pool Kt/V (spKt/V) using the following formula:where spKt/V is a single-pool Kt/V, R is the ratio of post dialysis to pre-dialysis blood urea nitrogen, t is time on dialysis in hours, UF is the amount of ultrafiltration in liters, and W is post-dialysis body weight in kilograms. The normalized protein equivalent of total nitrogen appearance (nPNA), also known as normalized protein catabolic rate, was calculated to estimate the daily protein intake by the following formula:where C is the pre-dialysis concentration of blood urea nitrogen in milligrams per deciliter.

Body mass index was calculated as the ratio of dry weight in kilograms (end-dialysis weight) to height squared (in square meters).

Laboratory parameters including serum hemoglobin, glucose, albumin, prealbumin, C-reactive protein, blood urea nitrogen, creatinine, uric acid, calcium, phosphorus, total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglyceride, iron, ferritin, transferrin, parathyroid hormone, cTnI, homocysteine, and tumor necrosis factor-alpha (TNF-α) were measured before the beginning of the hemodialysis session. Post-dialysis blood urea nitrogen concentration was used to calculate urea reduction ratio.

Cardiac troponin I was analyzed by a Microparticle Enzyme Immunoassay (Axsys System, Abbott Park, Illinois, USA) with a normal range of cTnI value from 0.02 to 2.3 ng/mL. The measuring range was 0.02–50 ng/mL. Cardiac troponin I was considered as elevated if the measurement was above the normal limit. Homocysteine was measured by fluorescent polarizing enzyme immunoassay (Axis Biomedicals ASA, Oslo, Norway) with a measuring range of 2–50 μmol/L, coefficient of variation: <10%, and normal laboratory range 5–15 μmol /L. Serum levels of TNF-α was measured with enzyme-linked immunosorbent assay method by using the Biosourche rat kit (Cat. no: KRC3011, Nivelles, Belgium). Other biochemical parameters were measured using standard laboratory methods.

Statistical Analysis

Statistical analysis was performed using SPSS 10.0 for Windows (SPSS Inc, Evanston, Illinois, USA). Results are considered statistically significant if two-tailed p is less than 0.05. Data are shown as mean ± standard deviation, median range and as a percentage (%). Comparisons between two groups were assessed by means of Mann Whitney-U test for continuous variables, as one of the groups was composed of 10 patients. Categorical variables were analyzed by the Fishers' exact test. Correlations were analyzed as either by Pearson (if two of the variables were normally distributed) or Spearman's correlation coefficients (if at least one of the variable was not normally distributed). A multiple linear regression analysis with backward elimination method was performed to identify possible independent predictors of cTnI. Variables included were age, sex, smoking status, diabetes mellitus, coronary artery disease, left ventricular hypertrophy, body mass index, uncontrolled blood pressure, nPNA, hemoglobin, albumin, HDL-C, LDL-C, triglyceride, homocysteine, C-reactive protein, and TNF-α. Because cTnI was not normally distributed, logarithmic conversion was performed before linear regression analysis.

RESULTS

In total, 113 hemodialysis patients (male/female, 66/47; mean age, 50.9 ± 11.9 years; mean hemodialysis duration, 99.7 ± 56.8 months) were included in the study. The etiologies for ESRD were glomerulonephritis in 21, hypertension 21, diabetes in 18, vesicourethral reflux/pyelonephritis in 16, nephrolithiasis in 10, polycystic kidney disease in 6, and amyloidosis in 4 patients. The etiology was unknown in 17 patients. Seventy-two (63.7%) patients were nonsmokers. Atherosclerosis was established in 46 patients (40.7%) and coronary artery disease in 38 (33.6%) patients. Peripheral artery disease and cerebrovascular diseases were present in 15 and 6 patients, respectively.

Cardiac troponin I concentrations were below the cutoff value (<2.3 ng/mL) in 103 (91.2%) patients (Group 1), whereas 10 (8.8%) patients had concentrations above the cutoff value (≥2.3 ng/mL) (Group 2). The comparative baseline characteristics, and demographic and clinical parameters of patients with normal and elevated cTnI are shown in Tables 1 and 2, respectively. When compared to Group 2, patients in Group 1 had higher hemoglobin and HDL-C levels and lower C-reactive protein and TNF-α levels (see Table 3).

Table 1  Comparative baseline characteristics of hemodialysis patients with normal and elevated cardiac troponin I levels

	Group 1: cTnI <2.3 ng/ml (n = 103)	Group 2: cTnI ≥2.3 ng/ml (n = 10)	p
Smoker/nonsmoker	38 / 65	3 / 7	0.745
Diabetes mellitus (P/A)	15 / 88	3 / 7	0.197
Atherosclerosis (P/A)	40 / 63	6 / 4	0.312
Coronary artery disease (P/A)	33 / 70	5 / 5	0.300
Peripheral artery disease (P/A)	12 / 91	3 / 7	0.128
Cerebrovascular disease (P/A)	5 / 98	1 / 9	0.434
Statin usage (P/A)	29 / 74	4 / 6	0.474
Left ventricular hypertrophy (P/A)	37 / 66	7 / 3	0.045
Anti-hypertensive medication (P/A)	52 / 51	3 /7	0.323

Abbreviations: cTnI = cardiac troponin I, P/A = present/absent.

Table 2  The comparative demographic and clinical parameters of hemodialysis patients with normal and elevated cardiac troponin I levels

	Group 1: cTnI <2.3 ng/ml (n = 103)	Group 2: cTnI ≥2.3 ng/ml (n = 10)	p
Age (years)	50.0 ± 11.9	60.2 ± 7.6	0.008
Dialysis duration (months)	96 (8–220)	93 (72–144)	0.852
Body mass index (kg/m^2)	23.5 ± 3.8	22.5 ± 2.2	0.447
Systolic blood pressure (mmHg)	134.9 ± 17.4	146.6 ± 14.2	0.056
Diastolic blood pressure (mmHg)	81.8 ± 13.7	88.5 ± 9.7	0.199
nPNA (g/kg/day)	1.29 ± 0.20	1.18 ± 0.14	0.094
spKt/V	1.35 ± 0.14	1.30 ± 0.05	0.390

Abbreviations: cTnI = cardiac troponin I, nPNA = normalized protein equivalent of total nitrogen appearance.

Table 3  Comparative laboratory parameters of hemodialysis patients with normal and elevated cardiac troponin I levels

	Group 1: cTnI <2.3 ng/ml (n = 103)	Group 2: cTnI≥2.3 ng/ml (n = 10)	p
Hemoglobin (g/L)	113.0 ± 12.0	100.0 ± 9.0	0.002
Glucose (mmol/L)	5.2 ± 1.1	5.4 ± 1.1	0.264
Albumin (g/L)	40.0 ± 3.8	38 ± 3.7	0.110
Prealbumin (mg/dl)	32.7 ± 11.2	25.9 ± 13.1	0.154
C-reactive protein (mg/L)	4.1 (0.37–23.7)	9.4 (2.2–19.5)	0.003
Blood urea nitrogen (mmol/L)	26.8 ± 5.4	25.0 ± 4.5	0.302
Creatitine (μmol/L)	813.3 ± 123.8	813.3 ± 79.6	0.887
Uric acid (μmol/L)	410.0 ± 65.4	386.7 ± 59.5	0.227
Calcium (mmol/L)	2.3 ± 0.2	2.3 ± 0.1	0.875
Phosphorus (mmol/L)	1.7 ± 0.4	1.4 ± 0.4	0.058
Total cholesterol (mmol/L)	4.70 ± 1.16	4.64 ± 1.06	0.952
HDL-C (mmol/L)	1.25 ± 0.28	0.93 ± 0.32	0.002
LDL-C (mmol/L)	2.60 ± 0.82	2.46 ± 0.67	0.653
Triglyceride (mmol/L)	1.8 (0.4–5.0)	2.0 (0.5–2.5)	0.701
Iron (μmol/L)	12.4 (1.4–39.4)	12.1 (6.1–25.2)	0.804
Ferritin (ng/ml)	260.5 (17–1342)	254 (123–695)	0.744
Transferrin (μg/dl)	164.7 ± 42.7	171.7 ± 46.5	0.952
Intact PTH (ng/ml)	281 (17–999)	313.5 (21.5–689)	0.864
Homocystein (μmol/l)	10.3 ± 7.5	10.0 ± 4.9	0.578
TNF-α (pg/ml)	12.8 ± 4.5	17.3 ± 4.5	0.005

Abbreviations: cTnI = cardiac troponin I, HDL-C = high-density lipoprotein cholesterol, LDL-C = low-density lipoprotein cholesterol, PTH = parathyroid hormone, TNF-α = tumor necrosis factor-alpha.

Cardiac troponin I was positively correlated with age (r = +0.223, p = 0.017), systolic blood pressure (r = +0.204, p = 0.03), and C-reactive protein (r = +0.425, p < 0.0001), and negatively correlated with albumin (r = −0.269, p = 0.004) and HDL-C (r = −0.295, p = 0.002) (data not shown). The regression graphics of cTnI with TNF-α, nPNA, and hemoglobin are shown in Figures 1–3, respectively. An exploration of factors that may be associated with high log cTnI by backward multiple linear regression analysis revealed that diabetes mellitus, left ventricular hypertrophy, uncontrolled blood pressure, nPNA, hemoglobin, and TNF-α were found to be independently associated with logcTnI (see Table 4).

Figure 1. The regression graphic between cardiac troponin I and tumor necrosis factor-alpha.

Figure 2. The regression graphic between cardiac troponin I and normalized protein equivalent of total nitrogen appearance.

Figure 3. The regression graphic between cardiac troponin I and hemoglobin.

Table 4  Backward multiple linear regression analysis of factors that may be associated with high log cTnI

	B	Beta	95% CI	p
Diabetes mellitus	+0.305	+0.160	0.046–0.565	0.021
Left ventricular hypertrophy	+0.356	+0.247	0.163–0.549	<0.0001
Uncontrolled blood pressure	+0.246	+0.170	0.047–0.444	0.016
nPNA	−0.834	−0.230	−1.328 – −0.340	0.001
Hemoglobin	−0.169	−0.302	−0.242 – −0.096	<0.0001
TNF-α	+0.076	+0.506	0.057–0.097	<0.0001

Abbreviations: cTnI = cardiac troponin I, CI = confidence interval, nPNA = normalized protein nitrogen appearance, TNF-α = tumor necrosis factor-alpha.

DISCUSSION

In ESRD patients, coronary artery disease is a major cause of death. Diagnosis of coronary artery disease in these patients is difficult, as many of these patients are asymptomatic and present with an abnormal electrocardiogram.^[3] As a marker of myocardial damage, cTnI is exclusively of cardiac origin^[4] and is accepted as an indicator of myocardial micro-necrosis.^[13] The elevation of cTnI was shown to be independent of the magnitude of renal insufficiency, and no significant difference has been found between pre- and post-dialysis values. Therefore, cTnI is considered to be a sensitive and a specific marker of myocardial injury in ESRD patients, even in the absence of clinical, electrocardiographic, or echocardiographic pathologies.^[3,4] In the present study, we searched for potential traditional and non-traditional risk factors in hemodialysis patients for cTnI elevation, which is considered to be a reliable marker of silent/subtle myocardial damage.

In our study, as a novel finding, we demonstrated significant correlation and independent relation between cTnI and TNF-α in clinically asymptomatic hemodialysis patients. Our findings are in accordance with the previous studies conducted in patients without renal failure.^[13,14] There is a dual relationship between myocardial cell death (either due to necrosis or apoptosis) and the release of TNF-α. On one hand, myocardial necrosis triggers the release of TNF-α both from inflammatory cells and from myocytes^[15,16]; on the other hand, TNF-α increases the rate of apoptotic cardiomyocyte death.^[17] In addition, during coronary arterial thrombosis and myocardial infarction, both TNF-α and troponin levels are increased.^[18] All of the above-mentioned evidence suggests that myocardial necrosis and the inflammatory marker, TNF-α, are interrelated. First of all, our findings confirm the interrelationship between cTnI and TNF-α levels in asymptomatic hemodialysis patients. Second, we have found that diabetes mellitus was independently associated with higher cTnI levels in hemodialysis patients. We do not know the mechanism behind the increased cTnI levels in diabetic ESRD patients. Previous studies have shown that diabetic patients had significantly higher levels cTnT levels compared to non-diabetics, and it was speculated that diabetic patients are at high risk for developing ischemic heart disease, which may be present but undiagnosed in this patient group.^[19,20] In contrast to our findings, in one study, cTnI (unlike cTnT) was not associated with diabetes mellitus in hemodialysis patients. The authors speculated that the observed disparities in their findings may be attributed to the detection or release of complex forms, and to different half lives, catabolism pathways, and transient membrane permeability changes that occur during myocardial cell injury.^[11]. We believe that more studies are needed to highlight these conflicting issues.

In our study, despite its relatively low prevalence (38.9%), there was an independent and positive relationship between left ventricular hypertrophy (detected by electrocardiography) and logcTnI levels. Ilıou et al. also found that 35.8% of patients had left ventricular hypertrophy detected by echocardiography, and, unlike cTnI, cTnT was related with left ventricular hypertrophy.^[11] Tun et al. also did not find any relationship between cTnI and cardiac hypertrophy.^[21] However, the relationship between left ventricular hypertrophy and myocardial ischemia is well established and based on several mechanisms (e.g., coronary artery disease, or decrease in the perfusion of subendocardial wall).^[22] Martin et al. demonstrated that renal failure patients with elevated cTnI showed cardiac abnormalities demonstrated by echocardiography, nuclear imaging, or arteriography.^[23] Additionally, it was previously speculated that in cardiac hypertrophy, more ultrastructural components were synthesized with increased synthesis of enzymes, as well as an increased number and size of cardiac fibers. This may lead to increased leakage of cardiac specific enzymes.^[20] Missov et al. found increased serum cTnI levels in patients with congestive heart failure and speculated that the membranes of injured myocytes may lose their integrity and allow the exposure of intracellular myocytes.^[24]

Uncontrolled blood pressure was associated with high logcTnI levels in our study. Elevated blood pressures in dialysis patients may induce left ventricular hypertrophy, which in turn may be responsible for higher logcTnI levels in these patients. However, in our study, although patients with left ventricular hypertrophy had higher pre-dialysis systolic and diastolic blood pressures, these differences did not have a statistical significance (data not shown). Thus, we speculate that higher blood pressure per se had an impact on elevation of cTnI, independent of its effect on left ventricular hypertrophy. It is known that essential hypertension is a well-known risk factor for myocardial infarction.^[25] It was previously speculated that there was a link between cardiac ischemia and systemic hypertension, although the exact mechanisms were not known.^[26] One of the postulated mechanisms is the impairment of coronary endothelial dysfunction by hypertension. Presence of inflammation^[27] and oxidative stress^[28] may play a role for the endothelial dysfunction. Thus, hypertension apart from inducing ventricular hypertrophy can also induce endothelial dysfunction by various mechanisms and induce cardiac damage.

We found that one of the independent factors for higher cTnI was the presence of lower hemoglobin levels. Ritz et al. suggested that anemia may represent an important risk factor for accelerated atherosclerosis in dialyzed patients.^[29] It is well-known that higher hematocrit values were associated with lower cardiac and all cause mortality in incident hemodialysis patients, even in those with cardiac disease.^[30] The exact role of anemia in increasing mortality in dialysis patients is not completely understood.^[31] One explanation may be that the association between anemia and mortality risk may be mediated through cardiovascular disease, malnutrition, and inflammation. In our study, hemoglobin was positively correlated with albumin (r = +0.327, p < 0.0001) and prealbumin (r = +0.170, p = 0.026), and was negatively correlated with C-reactive protein (r = −0.251, p = 0.001) and TNF-α (r = −0.287, p < 0.0001) (data not shown). Erythropoietin resistance index, defined as the average weekly erythropoietin dose divided by the average blood hemoglobin value,^[32] was significantly higher in the high cTnI group. Albumin (r = −0.334, p < 0.0001) and prealbumin (r = −0.230, p = 0.02) were negatively correlated with erythropoietin resistance index, whereas C-reactive protein (r = +0.232, p = 0.02) and TNF-α (r = +0.442, p < 0.0001) were positively correlated with erythropoietin resistance index (data not shown). All of our findings support the well-known notion that protein-energy malnutrition, inflammation, and refractory anemia, which are considered as non-traditional risk factors leading to cardiovascular disease, are closely related in patients with ESRD.^[33] We can speculate that anemia, which is closely related with inflammatory and nutritional markers, mediate its adverse effects through cardiovascular disease reflected as high cTnI levels in our study population. However, in our study, anemia was associated with higher cTnI levels independent of other inflammatory and nutritional markers. We suggest that, apart from association with protein-energy malnutrition and inflammation, anemia is directly involved in myocardial damage in hemodialysis patients at least in two ways; by limiting oxygen delivery in the presence of atherosclerotic lesions and by its effect on cardiac structure through the induction of left ventricular hypertrophy.^[34]

In dialysis patients, nPNA reflects daily protein intake. Even in adequately dialyzed patients, nPNA is correlated with mortality.^[35] Not only baseline nPNA, but changes in nPNA over time are associated with survival in hemodialysis patients, independent of other nutritional and inflammatory markers. It is suggested that protein intake, as detected by nPNA, might be an independent factor for persistent cardiovascular disease and mortality in dialysis patients, with mechanisms being unclear.^[36] Although we did not analyze mortality, our present findings first showed that there was an independent association between nPNA and higher cTnI levels. As already mentioned earlier, in asymptomatic hemodialysis patients, higher cTnI levels at the time of presentation might be related with subclinical myocardial damage and later be associated with major cardiovascular disease. By the light of our findings, we can speculate that even in asymptomatic HD patients, low nPNA levels associated with cardiac damage could result in higher all cause and cardiovascular mortality, even with adjustments after inflammation and other nutritional factors.

We found no independent relationship between homocysteine and higher cTnI levels. Conflicting findings are present about the relationship between homocysteine and cardiovascular disease in hemodialysis patients. In one five-year follow-up period of the study, Kumagai et al. found no relationship between homocysteine and atherosclerotic indices and cardiovascular events.^[37] Also, it was suggested that a worsening of atherosclerotic cardiovascular disease is mostly related to malnutrition-inflammation cachexia syndrome, a condition in which homocysteine levels are low instead of high.^[38]

Caglar et al. showed that silent myocardial ischemia detected by single photon emission computed tomography in patients undergoing hemodialysis had higher homocysteine levels.^[39] Thus, the results for our patients does not fulfill the terms of traditional epidemiology and paradoxical reverse epidemiology of homocysteine.

We are aware that our study has limitations, and our findings should be interpreted with caution. Because our study is not an experimental one, we cannot prove a cause-and-effect relationship. Due to the cross-sectional design and the fact that blood samples were taken only once, we were unable to examine impact of anemia, nPNA, and inflammation on cTnI levels over time. We did not examine cardiac structural changes and cardiac function (e.g., cardiac hypertrophy, ejection fraction) by more reliable methods like echocardiography. We detected cardiac hypertrophy by electrocardiogram, which is less sensitive than echocardiography in this respect. We also did not analyze cardiac atherosclerosis by conventional methods like angiography, and the presence of cardiovascular disease was based on history and medical records. In our study, only 8.8% of patients had cTnI levels above cutoff. At first glance, this seems to be a very small number. However, our findings are consistent with previous literature. Katerinitis et al. found that cTnI levels were persistently elevated in 8% of asymptomatic hemodialysis patients, which is very similar to our findings.^[40] Finally, it should be emphasized that the results obtained in this study may not be applicable to all patient population because of the impact of differences, in age, sex, race, comorbidity, and so on.

In conclusion, to the best of our knowledge, we first demonstrated that both traditional and non-traditional risk factors are independently associated with higher cTnI levels. We suggest that traditional and non-traditional risk factors are not only associated with manifest disease, but they are also risk factors for silent myocardial damage in asymptomatic hemodialysis patients.

DECLARATION OF INTEREST

The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.