The impact of low triiodothyronine levels on mortality is mediated by malnutrition and cardiac dysfunction in incident hemodialysis patients

in European Journal of Endocrinology
View More View Less
  • 1 Department of Internal Medicine, Severance Biomedical Science Institute, College of Medicine

(Correspondence should be addressed to S-W Kang who is now at Department of Internal Medicine, College of Medicine, Severance Biomedical Science Institute, Brain Korea 21, Yonsei University, 134 Shinchon-Dong, Seodaemoon-Gu, Seoul 120-752, Korea; Email: kswkidney@yuhs.ac)

Objective

Little is known about the impact of low triiodothyronine (T3) levels on mortality in end-stage renal disease (ESRD) patients starting hemodialysis (HD) and whether this impact is mediated by malnutrition, inflammation, or cardiac dysfunction.

Design and methods

A prospective cohort of 471 incident HD patients from 36 dialysis centers within the Clinical Research Center for ESRD in Korea was selected for this study. Based on the median value of T3, patients were divided into ‘higher’ and ‘lower’ groups, and all-cause and cardiovascular (CV) mortality rates were compared. In addition, associations between T3 levels and various nutritional, inflammatory, and echocardiographic parameters were determined.

Results

Compared with those in the ‘higher’ T3 group, albumin, cholesterol, and triglyceride levels, lean body mass estimated by creatinine kinetics (LBM-Cr), and normalized protein catabolic rate (nPCR) were significantly lower in patients with ‘lower’ T3 levels. The ‘lower’ T3 group also had a higher left ventricular mass index (LVMI) and a lower ejection fraction (EF). Furthermore, correlation analysis revealed significant associations between T3 levels and nutritional and echocardiographic parameters. All-cause and CV mortality rates were significantly higher in patients with ‘lower’ T3 levels than in the ‘higher’ T3 group (113.4 vs 18.2 events per 1000 patient-years, P<0.001, and 49.8 vs 9.1 events per 1000 patient-years, P=0.001, respectively). The Kaplan–Meier analysis also showed significantly worse cumulative survival rates in the ‘lower’ T3 group (P<0.001). In the Cox regression analysis, low T3 was an independent predictor of all-cause mortality even after adjusting for traditional risk factors (hazard ratio=3.76, P=0.021). However, the significant impact of low T3 on all-cause mortality disappeared when LBM-Cr, nPCR, LVMI, or EF were incorporated into the models.

Conclusion

Low T3 has an impact on all-cause mortality in incident HD patients, partly via malnutrition and cardiac dysfunction.

Abstract

Objective

Little is known about the impact of low triiodothyronine (T3) levels on mortality in end-stage renal disease (ESRD) patients starting hemodialysis (HD) and whether this impact is mediated by malnutrition, inflammation, or cardiac dysfunction.

Design and methods

A prospective cohort of 471 incident HD patients from 36 dialysis centers within the Clinical Research Center for ESRD in Korea was selected for this study. Based on the median value of T3, patients were divided into ‘higher’ and ‘lower’ groups, and all-cause and cardiovascular (CV) mortality rates were compared. In addition, associations between T3 levels and various nutritional, inflammatory, and echocardiographic parameters were determined.

Results

Compared with those in the ‘higher’ T3 group, albumin, cholesterol, and triglyceride levels, lean body mass estimated by creatinine kinetics (LBM-Cr), and normalized protein catabolic rate (nPCR) were significantly lower in patients with ‘lower’ T3 levels. The ‘lower’ T3 group also had a higher left ventricular mass index (LVMI) and a lower ejection fraction (EF). Furthermore, correlation analysis revealed significant associations between T3 levels and nutritional and echocardiographic parameters. All-cause and CV mortality rates were significantly higher in patients with ‘lower’ T3 levels than in the ‘higher’ T3 group (113.4 vs 18.2 events per 1000 patient-years, P<0.001, and 49.8 vs 9.1 events per 1000 patient-years, P=0.001, respectively). The Kaplan–Meier analysis also showed significantly worse cumulative survival rates in the ‘lower’ T3 group (P<0.001). In the Cox regression analysis, low T3 was an independent predictor of all-cause mortality even after adjusting for traditional risk factors (hazard ratio=3.76, P=0.021). However, the significant impact of low T3 on all-cause mortality disappeared when LBM-Cr, nPCR, LVMI, or EF were incorporated into the models.

Conclusion

Low T3 has an impact on all-cause mortality in incident HD patients, partly via malnutrition and cardiac dysfunction.

Introduction

Thyroid hormone has numerous effects on the kidney, playing an important role in renal development and growth as well as in sodium and water homeostasis. Renal plasma flow (RPF) and glomerular filtration rate (GFR) are also influenced by thyroid hormone. As a result, reductions in RPF and GFR, an increase in serum creatinine concentrations, and hyponatremia are frequently observed in primary hypothyroidism patients, and these renal derangements can be nearly normalized by thyroid hormone replacement therapy (1, 2, 3).

Conversely, the kidney is involved in the metabolism and excretion of thyroid hormone; thus, renal failure can lead to significant changes in thyroid function. Goiter, thyroid nodules, reduced total and free 3,5,3'-triiodo-L-thyronine (fT3) and thyroxine (T4) levels, and subclinical hypothyroidism are not uncommon in patients with chronic kidney disease (CKD). Among these abnormalities, reduced serum T3 concentration (low T3) is the most frequently observed alteration of thyroid hormone in CKD patients (1, 4). A recent study has shown that there is a significant correlation between serum T3 levels and estimated GFR (eGFR) and that the prevalence of low T3 gradually increases with worsening CKD stage (5). In addition, some previous studies have demonstrated that up to 70% of stage 5 CKD patients have low T3 (6). This alteration in CKD patients has mainly been attributed to impaired TSH response to thyrotropin-releasing hormone and decreased peripheral biosynthesis of T3 from T4 due to uremia and has been considered to be merely an adaptive process to preserve energy (1, 7). Recently, however, low T3 has also been suggested to be associated with metabolic acidosis, systemic inflammation, endothelial dysfunction, atherosclerosis, and cardiomyopathy in patients with end-stage renal disease (ESRD) (8, 9, 10, 11, 12).

Accumulating evidence has indicated that low T3 is associated with clinical outcomes in patients with various chronic diseases, including congestive heart failure, respiratory failure, and liver cirrhosis (13, 14, 15, 16). Since low T3 is the most common disturbance of thyroid hormone levels in ESRD patients, the adverse impact of low T3 on patient survival has been widely explored in patients on maintenance hemodialysis (HD) and peritoneal dialysis (PD) (17, 18). Furthermore, low pre-transplant serum fT3 levels have been revealed to be a significant predictor of subsequent graft failure in kidney transplant recipients (19). In spite of a large number of previous investigations on the prognostic value of low T3 for all-cause and cardiovascular (CV) mortality in ESRD patients, the majority of the study subjects were prevalent HD or PD patients and echocardiographic findings were not included in the analysis in most of these studies. In addition, compared with fT3 concentrations, total T3 levels were more commonly determined in many hospitals and were less influenced by the way of measurement or medications such as heparin (20, 21, 22), but clinical outcomes were more frequently defined based on serum or plasma fT3 concentrations than on total T3 concentrations.

Therefore, in this study, we determined whether low serum total T3 levels were a significant predictor of mortality in Korean incident HD patients from the Clinical Research Center for ESRD (CRC for ESRD) cohort. Moreover, the associations between T3 concentrations and metabolic acidosis, systemic inflammation, malnutrition, and echocardiographic parameters were investigated in these patients.

Subjects and methods

Subjects

Initial recruitment for this prospective observational multi-center study involved all ESRD patients starting HD between August 1, 2008, and October 31, 2012, at 36 centers of the CRC for ESRD in Korea. Within 7 days of the initial dialysis, the patients provided written informed consent and were enrolled into the study cohort. Among these patients, we excluded those who were aged <18 years, had a history of PD or kidney transplantation prior to HD, were on thyroid hormone replacement or suppressive therapy due to overt thyroid disease, were on medications affecting thyroid hormone levels, such as amiodarone, glucocorticoids, and lithium, had underlying active malignancy, or died within 3 months of the commencement of HD. Patients who had insufficient baseline data, including thyroid function test results, were also excluded from the study. Ultimately, a total of 471 incident HD patients were included in the final analysis.

This study was carried out in accordance with the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board at each participating center.

Data collection

Demographic and clinical data were recorded at the time of study entry, including age, gender, BMI calculated as weight/height2, primary renal disease, comorbidities, and medications. Coronary arterial disease (CAD) was defined as a history of angioplasty, coronary artery bypass grafts, myocardial infarction, or angina, while peripheral arterial disease (PAD) was defined as a history of claudication, ischemic limb loss and/or ulceration, or peripheral revascularization procedure. The composite of CAD, congestive heart failure of any cause ischemic, valvular disease related or amyloid deposition related to uremia, and arrhythmia was designated as cardiac disease, while the composite of cardiac disease, PAD, and cerebrovascular disease, transient ischemic attack, ischemia/infarction, or bleeding as cardiovascular disease (CVD). Laboratory data were measured using pre-dialysis fasting blood samples collected on the day of a mid-week dialysis session, when the patients were considered to be clinically stable and to be in euvolemic state. These laboratory data included data on hemoglobin (Hb), white blood cell (WBC) count, blood urea nitrogen, creatinine, calcium, phosphorus, intact parathyroid hormone, albumin, total cholesterol, triglyceride, alkaline phosphatase, HbA1c, sodium, potassium, bicarbonate, serum iron, ferritin, N-terminal pro-B-type natriuretic peptide (NT-proBNP), cardiac troponin T (cTnT), and high-sensitivity C-reactive protein (hs-CRP) levels. During follow-up, these data were measured every 3 months according to the K/DOQI guideline (http://www.kidney.org, June 26, 2013). eGFR was calculated using the four-variable Modification of Diet in Renal Disease (MDRD) Study and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) Study equation (23). In addition, 24-h urine sample collection was executed to determine 24-h urinary protein, urea, and creatinine excretion. Nutritional status was also evaluated using subjective global assessment (SGA) score (24), lean body mass estimated by creatinine kinetics (LBM-Cr) (25), and normalized protein catabolic rate (nPCR) (26). Thyroid function test was performed 3 months after the initiation of HD and annually thereafter based on the consideration of the economic issue and the opinion of the clinician who participated in this prospective observational study. Serum T3, free T4 (fT4), and TSH levels were determined using a chemiluminescent microparticle immunoassay on the Architect-i2000SR analyzer (Abbott Laboratories), and the reference ranges were 0.58–1.59 ng/ml, 0.70–1.48 ng/dl, and 0.35–4.94 μIU/ml respectively. Since there was an extremely wide variation in the biochemical data at 0 month due to diverse patients' conditions to start HD, we used the laboratory data collected at 3 months after HD commencement, when correction of uremia, anemia, and acidosis was considered to be sufficiently achieved.

Echocardiography was performed on a non-dialysis day, close to the time of discharge, based on the imaging protocol recommended by the American Society of Echocardiography. Left atrial dimension (LAD) was assessed at end-ventricular systole at the level of the aortic valve according to the leading-edge-to-leading-edge convention. Left ventricular (LV) mass was determined using the method described by Steiber et al. (24), and the LV mass index (LVMI) was calculated by dividing LV mass by body surface area. LV systolic function was defined by LV ejection fraction (LVEF) using a modified biplane Simpson method from the apical two-chamber and four-chamber views. Multiple reproducibility, inter-reader reliability, intra-reader reliability, and reader drift analyses were performed at a core echocardiography laboratory (Kyungpook National University, Daegu, Korea) on a random sample of 3% of the entire cohort each year. The intra-class correlation coefficients for the echocardiographic measures were 0.773 for LAD, 0.745 for LVMI, and 0.842 for LVEF.

Outcome measures

For this study, data on all mortality events were retrieved from the database and carefully reviewed. The primary and secondary endpoints were all-cause mortality and CV mortality respectively. CV mortality was considered death from myocardial infarction or ischemia, congestive heart failure, pulmonary edema, cerebral hemorrhage, or cerebrovascular disorder.

Statistical analysis

Statistical analysis was performed using SPSS for Windows, version 18.0 (SPSS, Inc.). Continuous variables are expressed as mean±s.d. or median (interquartile range) and categorical variables as a number (percentage). Normality of the distribution was ascertained using the Shapiro–Wilk test. The patients were dichotomized into ‘higher’ and ‘lower’ T3 groups based on the median values of serum T3 concentrations. Patient demographics, clinical characteristics, and laboratory findings were compared between the two groups using Student's t-test or Mann–Whitney U test for continuous variables and the χ2 test for categorical variables. Because of the log-normal distributions of NT-proBNP, cTnT, and hs-CRP concentrations, natural log values (Ln) were used in the analysis. The associations of serum T3 levels with other variables were determined using Pearson's correlation analysis. A multivariate linear regression analysis was conducted to identify the independent determinants of T3 concentrations.

Cumulative survival curves for all-cause and CV mortality were created using the Kaplan–Meier method, and between-group survival was compared using a log-rank test. The independent prognostic power of serum T3 levels for mortality was ascertained using multivariate Cox proportional-hazards regression analysis, which included only the variables of P values <0.10 in the univariate analysis. The positive predictive value of T3 concentrations for mortality was also analyzed using the receiver operating characteristic (ROC) curve analysis with calculated area under the ROC curve (AUC). P values <0.05 were considered statistically significant.

Results

Baseline characteristics

The baseline demographic and clinical characteristics of the patients are given in Table 1. The mean age was 56.9±14.9 years and 57.3% of the patients were male. The most common cause of ESRD was diabetes (DM, 53.3%), followed by hypertension (14.4%). The median values of serum T3, fT4, and TSH concentrations were 0.8 ng/ml, 1.1 ng/dl, and 1.9 μIU/ml respectively.

Table 1

Baseline demographic and clinical characteristics of the study subjects.

VariablesTotal (n=471)T3<0.8 ng/ml (n=228)T3≥0.8 ng/ml (n=243)P
Age (years)56.85±14.9458.63±14.9655.19±14.760.012
Sex (male)270 (57.3%)116 (50.9%)154 (63.4%)0.006
BMI (kg/m2)23.14±3.6922.71±3.8923.55±3.480.065
Systolic BP (mmHg)143.0±23.4142.7±22.2143.3±24.50.814
Diastolic BP (mmHg)77.6±14.076.7±14.378.5±13.60.159
Smoking status0.049
 Current smoker54 (11.5%)30 (13.2%)24 (9.9%)
 Ex-smoker138 (29.3%)55 (24.1%)83 (34.2%)
 Non-smoker279 (59.2%)143 (62.7%)136 (56.0%)
Primary cause of renal disease0.636
 Diabetes251 (53.3%)128 (56.1%)123 (50.6%)
 Hypertension68 (14.4%)35 (15.4%)33 (13.6%)
 Chronic GN69 (14.6%)29 (12.7%)40 (16.5%)
 Othersa50 (10.6%)22 (9.6%)28 (11.5%)
 Unknown33 (7.0%)15 (6.6%)18 (7.4%)
Comorbid disease
 Chronic lung disease47 (10.0%)25 (11.0%)22 (9.1%)0.497
 Coronary arterial disease72 (15.3%)40 (17.5%)32 (13.2%)0.186
 Peripheral arterial disease47 (10.0%)27 (11.8%)20 (8.2%)0.195
 Cerebrovascular disease62 (13.2%)36 (15.8%)26 (10.7%)0.103
 Congestive heart failure73 (15.5%)42 (18.4%)31 (12.8%)0.093
 Arrhythmia13 (2.8%)7 (3.1%)6 (2.5%)0.690
 Diabetes260 (55.2%)131 (57.5%)129 (53.1%)0.366
 Connective tissue disease50 (10.6%)24 (10.5%)26 (10.7%)0.941
 Ulcer42 (8.9%)21 (9.2%)21 (8.6%)0.827
 Liver disease47 (10.0%)19 (8.3%)26 (10.7%)0.383
 Cardiac diseaseb121 (25.7%)68 (29.8%)53 (21.8%)0.047
 CVDc172 (36.5%)96 (42.1%)76 (31.3%)0.015
Modified CCI5.16±2.455.44±2.524.90±2.360.017
Medications
 RAS blockers241 (51.2%)123 (53.9%)118 (48.6%)0.242
 CCB268 (56.9%)127 (55.7%)141 (58.0%)0.611
 Beta blocker251 (53.3%)114 (50.0%)137 (56.4%)0.166
 Aspirin106 (22.5%)58 (25.4%)48 (19.8%)0.140
 Clopidogrel34 (7.2%)22 (9.6%)12 (4.9%)0.048
 Vitamin D64 (13.6%)35 (15.4%)29 (11.9%)0.280
 ESA159 (33.8%)69 (30.3%)90 (37.0%)0.120

BP, blood pressure; GN, glomerulonephritis; CVD, cardiovascular disease; CCI, Charlson comorbidity index; RAS, renin–angiotensin system; CCB, calcium channel blocker; ESA, erythropoietin-stimulating agent.

Other causes of renal failure include polycystic kidney disease, interstitial nephritis, acquired obstructive uropathy, chronic pyelonephritis, reflux nephropathy, and amyloidosis.

Composite of coronary arterial disease, congestive heart failure, and arrhythmia.

Composite of coronary arterial disease, congestive heart failure, arrhythmia, peripheral arterial disease, and cerebrovascular disease.

The demographic and clinical characteristics were compared between the dichotomized two groups based on the median levels of T3 (0.8 ng/ml). The mean age, the proportion of female patients, and modified Charlson comorbidity index (CCI) were significantly higher in the ‘lower’ T3 group than in the ‘higher’ T3 group (P<0.05 or <0.01). In addition, CVDs were significantly more prevalent in the ‘lower’ T3 group (P=0.015). There was no significant difference in medication history between the two groups, except for the use of clopidogrel, which was prescribed more frequently in the ‘lower’ T3 group than in the ‘higher’ T3 group (P<0.048).

With regard to the laboratory and echocardiographic findings, patients with ‘lower’ T3 levels demonstrated blood glucose and serum uric acid levels that were significantly higher, while Hb, serum albumin, total cholesterol, triglyceride, and bicarbonate concentrations were significantly lower than those of the ‘higher’ T3 group (P<0.05–P<0.001). However, there were no differences in serum calcium, phosphorus, NT-proBNP, cTnT, and hs-CRP levels and eGFR between the two groups. The proportion of malnourished patients assessed by SGA was significantly higher (P=0.02) and LBM-Cr and nPCR were significantly lower in the ‘lower’ T3 group than in the ‘higher’ T3 group (P<0.05). Among 379 patients (192 in the ‘lower’ T3 group and 187 in the ‘higher’ T3 group) who underwent echocardiography, the ‘lower’ T3 group had a significantly higher LVMI (P=0.002) and a lower LVEF (P=0.012) than the ‘higher’ T3 group (Table 2).

Table 2

Laboratory and echocardiographic findings of the study subjects.

VariablesTotal (n=471)T3<0.8 ng/ml (n=228)T3≥0.8 ng/ml (n=243)P
Laboratory data at 3 months
 Hemoglobin (g/dl)10.84±1.4110.58±1.3211.01±1.450.001
 WBC (/μl)5883.4±2996.45911.4±3040.15282.5±3106.50.032
 Calcium (mg/dl)8.67±0.818.69±0.878.63±0.760.731
 Phosphorus (mg/dl)4.70±1.444.71±1.434.70±1.450.909
 Uric acid (mg/dl)6.91±1.967.10±1.936.74±1.980.043
 iPTH (pg/ml)182.59±54.27180.45±47.18184.57±60.930.789
 Glucose (mg/dl)124.17±44.24137.77±55.46111.58±40.730.003
 HbA1c (%) 6.10±1.416.14±1.506.06±1.290.453
 Protein (g/dl)6.64±0.736.49±0.716.76±0.73<0.001
 Albumin (g/dl)3.76±0.523.61±0.523.89±0.47<0.001
 BUN (mg/dl)54.89±21.3856.56±19.6453.27±22.920.258
 Creatinine (mg/dl)7.75±3.167.84±3.197.67±2.980.293
 GFR-MDRD (ml/min per 1.73 m2)6.92±3.716.86±2.946.99±4.380.187
 GFR-EPI (ml/min per 1.73 m2)7.14±3.747.04±2.957.23±4.420.218
 Total cholesterol (mg/dl)153.20±36.69147.05±37.40158.99±35.710.019
 Triglyceride (mg/dl)126.01±46.49116.77±41.84134.12±54.700.012
 Sodium (mmol/l)138.11±4.17137.83±4.77138.35±3.500.067
 Potassium (mmol/l)4.73±0.834.74±0.794.72±0.870.726
 Bicarbonate (mmol/l)22.47±4.4121.80±4.3923.12±4.350.007
 hs-CRP (mg/dl)0.14 (0.04–0.55)0.16 (0.04–0.74)0.13 (0.04–0.50)0.167
 NT-proBNP (pg/ml)7377.0 (2074.0–26 328.0)8613.0 (2124.0–29 821.5)5385.5 (1940.3–18 780.0)0.107
 cTnT (ng/ml)0.054 (0.028–0.113)0.060 (0.030–0.126)0.047 (0.026–0.093)0.110
 T3 (ng/ml)0.80±0.280.58±0.161.00±0.19<0.001
 Free T4 (ng/dl)1.10±0.261.08±0.241.12±0.280.084
 TSH (μIU/ml)1.90 (1.24–3.32)1.90 (1.20–3.29)1.94 (1.24–3.35)0.899
24-h urine study
 Urea (mg/dl)247.6 (158.9–451.7)238.0 (158.1–353.0)261.5 (163.8–970.0)0.129
 Creatinine (mg/dl)68.2 (42.6–181.4)64.6 (39.3–114.9)76.9 (44.9–327.2)0.017
 Protein (mg/day)922.4 (349.2–2109.1)772.2 (348.4–1767.4)1056.0 (353.7–2461.7)0.124
 Urine volume (ml/day)780.0 (460.0–1100.0)700.0 (400.0–1022.5)800.0 (490.0–1150.0)0.118
Nutritional markers
 SGA >1 (malnutrition)147 (31.2%)83 (36.4%)64 (26.3%)0.020
 LBM-Cr44.89±6.0844.20±6.0545.72±6.050.033
 nPCR0.94±0.390.86±0.341.05±0.460.021
 nPCR ≥1.0118 (25.1%)52 (22.8%)66 (27.2%)0.007
Baseline echocardiographic parameters
 LAD (cm)4.30±0.744.28±0.744.33±0.740.476
 LVMI (g/m2)144.05±49.91151.71±56.54136.04±40.520.002
 LVEF (%)58.66±11.3157.22±12.4460.13±9.830.012

WBC, white blood cell; iPTH, intact parathyroid hormone; BUN, blood urea nitrogen; GFR, glomerular filtration rate; MDRD, Modification of Diet in Renal Disease; EPI, Chronic Kidney Disease Epidemiology Collaboration equation; hs-CRP, high-sensitivity C-reactive protein; NT-proBNP, N-terminal pro-B-type natriuretic peptide; cTnT, cardiac troponin T; SGA, subjective global assessment; LBM-Cr, lean body mass estimated by creatinine kinetics; nPCR, normalized protein catabolic rate; LAD, left atrial diameter; LVMI, left ventricular mass index; LVEF, left ventricular ejection fraction.

Associations of T3 levels with malnutrition and inflammation markers and echocardiographic findings

Pearson's correlation analysis revealed that serum T3 concentrations had significant inverse correlations with age (r=−0.156, P=0.001), serum glucose levels (r=−0.180, P<0.001), WBC counts (r=−0.114, P=0.016), and Ln hs-CRP (r=−0.150, P=0.003). In contrast, there were significant positive associations between serum T3 concentrations and serum bicarbonate (r=0.234, P<0.001), albumin (r=0.354, P<0.001), total cholesterol (r=0.179, P=0.001), and triglyceride (r=0.193, P<0.001) levels, LBM-Cr (r=0.174, P=0.003), and nPCR (r=0.223, P<0.001). Furthermore, serum T3 concentrations had a significant inverse correlation with LVMI (r=−0.157, P=0.002) and a significant positive relationship with LVEF (r=0.168, P=0.001) (Table 3 and Fig. 1).

Table 3

Correlations of serum T3 levels with various nutritional, inflammatory, and echocardiographic variables.

T3 (ng/ml)
VariablesrP
Age (years)−0.1560.001
Sex (male)−0.162<0.001
Glucose−0.180<0.001
HbA1c (%)−0.0420.432
Bicarbonate (mmol/l)0.234<0.001
Nutritional markers
 BMI (kg/m2)0.0760.104
 Albumin (g/dl)0.354<0.001
 Total cholesterol0.1790.001
 Triglyceride0.193<0.001
 LBM-Cr0.1740.003
 nPCR0.223<0.001
Inflammatory markers
 White blood cell (/μl)−0.1140.016
 Ln hs-CRP−0.1500.003
Echocardiographic parameters
 LAD (cm)0.0010.978
 LVMI (g/m2)−0.1570.002
 LVEF (%)0.1680.001

LBM-Cr, lean body mass estimated by creatinine kinetics; nPCR, normalized protein catabolic rate; Ln hs-CRP, log-transformed high-sensitivity C-reactive protein; LAD, left atrial diameter; LVMI, left ventricular mass index; LVEF, left ventricular ejection fraction.

Figure 1
Figure 1

The correlation of serum T3 levels with various nutritional, inflammatory, and echocardiographic variables. Serum T3 concentrations had significant positive associations with (a) serum albumin levels, (b) LBM-Cr, (c) nPCR, and (d) LVEF, while there were significant inverse correlations of T3 with (e) LVMI and (f) Ln hs-CRP. LBM-Cr, lean body mass estimated by creatinine kinetics; nPCR, normalized protein catabolic rate; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; Ln hs-CRP, log-transformed high-sensitivity C-reactive protein.

Citation: European Journal of Endocrinology 169, 4; 10.1530/EJE-13-0540

Furthermore, a significant association of serum T3 levels with serum albumin (r=0.244, P=0.002) and bicarbonate (r=0.348, P<0.001) concentrations still persisted in 163 patients even at 12 months (Supplementary Figure 1, see section on supplementary data given at the end of this article).

Determinants of serum T3 levels

The multivariate linear regression analysis, including variables with P values <0.1 in the univariate analysis, demonstrated that age, serum albumin, bicarbonate, triglyceride, and hs-CRP concentrations, and nPCR were significant independent determinants of serum T3 levels (Table 4).

Table 4

Determinants of serum T3 concentrations.

T3
VariablesRegression coefficientP
Age (years)−0.0030.017
Sex (male vs female)0.0250.448
Underlying cardiovascular disease0.0270.424
Hemoglobin (g/dl)0.0090.489
Glucose (mg/dl)−0.0220.749
Albumin (g/dl)0.0680.047
Total cholesterol (mg/dl)0.0030.983
Triglyceride (mg/dl)0.1790.013
Bicarbonate (mmol/l)0.0100.006
SGA >1 (vs SGA ≤1)−0.0330.354
nPCR ≥1.0 (vs nPCR <1.0)0.1070.001
hs-CRP ≥5 mg/dl (vs hs-CRP <5 mg/dl)−0.1920.019

SGA, subjective global assessment; nPCR, normalized protein catabolic rate; hs-CRP, high-sensitivity C-reactive protein.

Clinical outcomes according to serum T3 concentrations

During a mean follow-up duration of 24.2±17.7 months, 49 patients (10.4%) died. Among them, 22 patients (44.9%) died from CV causes and 12 (24.5%) from infection. All-cause, CV, and infection-related mortality rates were significantly higher in patients with ‘lower’ T3 levels than in the ‘higher’ T3 group (all-cause mortality: 113.4 vs 18.2 events per 1000 patient-years, P<0.001; CV mortality: 49.8 vs 9.1 events per 1000 patient-years, P=0.001; and infection-related mortality: 27.7 vs 4.5 events per 1000 patient-years, P=0.014) (Table 5). The Kaplan–Meier analysis also indicated that the ‘lower’ T3 group had significantly worse all-cause and CV cumulative survival rates compared with the ‘higher’ T3 group (P<0.001 and P=0.001 respectively) (Fig. 2). Overall hospitalization rates, as well as composite death and hospitalization rates, were also significantly higher in patients with ‘lower’ T3 levels than in the ‘higher’ T3 group (Table 5).

Table 5

Comparisons of clinical outcomes according to serum T3 levels.

T3<0.8 ng/ml (n=228)T3≥0.8 ng/ml (n=243)
n (%)Rates (per 1000 patient-years)n (%)Rates (per 1000 patient-years)P
Mortality
 All-cause41 (18.0%)113.368 (3.3%)18.15<0.001
 CV18 (7.9%)49.774 (1.6%)9.080.001
 Infection10 (4.4%)27.652 (0.8%)4.540.014
Hospitalization
 All-cause102 (44.7%)282.0775 (30.9%)170.160.002
 CV35 (15.4%)96.7724 (9.9%)54.450.073
 Infection29 (12.7%)80.1814 (5.8%)31.760.009
Compositea
 All-cause132 (57.9%)364.9679 (32.5%)179.24<0.001
 CV48 (21.1%)132.7124 (9.9%)54.450.001
 Infection33 (14.5%)91.2414 (5.8%)31.760.002

CV, cardiovascular.

Composite: composite death and hospitalization.

Figure 2
Figure 2

Kaplan–Meier plots for all-cause and cardiovascular survival. Compared with patients with ‘higher’ T3, the ‘lower’ T3 group showed significantly worse (a) all-cause and (b) cardiovascular cumulative survival rates.

Citation: European Journal of Endocrinology 169, 4; 10.1530/EJE-13-0540

On the other hand, serum T3 levels provided significant positive predictive values for all-cause mortality before (AUC=0.674, P=0.005) and even after adjusting for BMI, smoking status, modified CCI, aspirin and clopidogrel usage, serum HbA1c, albumin, and potassium concentrations, eGFR by CKD-EPI, Ln hs-CRP, and SGA score (AUC=0.878, P<0.001) (Fig. 3). There was also a significant positive predictive value of T3 levels for CV death (AUC=0.702, P=0.007) (data not shown).

Figure 3
Figure 3

Receiver operating characteristic curves of T3 for all-cause mortality. Serum T3 concentrations provided significant positive predictive values for all-cause mortality (a) in an unadjusted model (AUC=0.674) and (b) in a model adjusted for BMI, smoking status, modified CCI, aspirin and clopidogrel usage, serum HbA1c, albumin, and potassium concentrations, eGFR by CKD-EPI, Ln hs-CRP, and SGA score (AUC=0.878). CCI, Charlson comorbidity index; eGFR, estimated glomerular filtration rate; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; Ln hs-CRP, log-transformed high-sensitivity C-reactive protein; SGA, subjective global assessment.

Citation: European Journal of Endocrinology 169, 4; 10.1530/EJE-13-0540

Low T3 as a predictor of all-cause mortality

The univariate Cox proportional-hazards regression analysis revealed that BMI, smoking, modified CCI, aspirin and clopidogrel usage, eGFR by CKD-EPI, serum HbA1c, albumin, and potassium concentrations, Ln hs-CRP, LBM-Cr, nPCR <1.0, LVMI, LVEF, and low T3 were significant risk factors for all-cause mortality (Supplementary Table 1, see section on supplementary data given at the end of this article). In multivariate Cox models, low T3 remained as a significant independent predictor of all-cause mortality (hazard ratio (HR)=3.758, 95% CI=1.055–13.382, P=0.021), even after adjusting for BMI, smoking status, modified CCI, aspirin and clopidogrel usage, eGFR by CKD-EPI, serum HbA1c, albumin, and potassium levels, Ln hs-CRP, and SGA score. However, when LBM-Cr, nPCR, LVMI, or LVEF were included in the successive models for further adjustment, the significant impact of low T3 on all-cause mortality disappeared (Table 6).

Table 6

Multivariate Cox proportional-hazards regression analysis for all-cause mortality.

Serum T3
HR95% CIP
Model 13.7581.055–13.3820.021
Model 27.7150.976–60.9870.053
Model 37.6920.973–60.8160.052
Model 46.8600.839–56.0960.072
Model 58.0080.971–66.0460.053

Model 1: adjusted for BMI, smoking status, modified Charlson comorbidity index, aspirin and clopidogrel usage, glomerular filtration rate estimated by Chronic Kidney Disease Epidemiology Collaboration equation, serum HbA1c, albumin, and potassium concentrations, log-transformed high-sensitivity C-reactive protein, and subjective global assessment score; Model 2: Model 1+lean body mass estimated by creatinine kinetics; Model 3: Model 1+normalized protein catabolic rate ≥1.0 (vs <1.0); Model 4: Model 1+left ventricular mass index; Model 5: Model 1+left ventricular ejection fraction.

In another analysis, we divided the patients into ‘subnormal’ and ‘normal’ T3 groups based on the lowest normal value of T3 (0.58 ng/ml) and determined the impact of ‘subnormal’ T3 on the clinical outcome. Similarly, ‘subnormal’ T3 was a significant risk factor for all-cause mortality (HR=5.943, 95% CI=1.442–24.492, P=0.014), but its statistical power disappeared when LBM-Cr, nPCR, LVMI, or LVEF were further incorporated into the models (Supplementary Tables 2 and 3, see section on supplementary data given at the end of this article).

Furthermore, low T3 was not a significant independent predictor of CV mortality after adjusting for smoking status, DM, underlying CVDs, eGFR by CKD-EPI, serum albumin and potassium concentrations, Ln hs-CRP, and SGA score (HR=2.735, 95% CI=0.917–11.395, P=0.086) (data not shown).

Discussion

Thyroid hormone abnormalities are common in CKD patients. Among these, low T3 is the most frequently observed alteration of thyroid hormone in patients with CKD. Recently, accumulating evidence has indicated that low T3 may act as a prognostic factor for adverse clinical outcomes in prevalent ESRD patients rather than as an innocent bystander (17, 18). In this study, we demonstrated that low T3 is a significant independent risk factor for all-cause mortality in incident HD patients. Moreover, it seems that the impact of low T3 on mortality may be due, in part, to malnutrition and cardiac dysfunction.

Low T3 has been reported to be present in up to 70% of patients with ESRD (5, 6). It has been proposed that multiple mechanisms contribute to this high prevalence of low T3, including systemic inflammation, malnutrition, and metabolic acidosis (8, 9). Zoccali et al. (17) showed that fT3 was a significant independent predictor of mortality even after adjusting for a series of traditional and non-traditional risk factors including interleukin-6 (IL6) in a multivariate analysis, but that this inflammatory cytokine predicted mortality only in a model excluding fT3. These findings suggest that fT3 and inflammation are involved in the same pathogenic pathway leading to death in ESRD patients and that fT3 mediates part of the adverse effects of inflammation in these patients. Moreover, similar findings have been demonstrated in prevalent PD patients (18). We also found that plasma T3 levels were significantly associated with hs-CRP concentrations and that hs-CRP was a significant independent determinant of T3. In contrast to a majority of previous studies, however, there was no difference in hs-CRP levels between the ‘lower’ and ‘higher’ T3 groups in the present study. It is surmised that this discrepancy may partly be due to differing ethnicities and dialysis duration (prevalent vs incident). Further investigation, including various inflammatory markers in a large, ethnically homogenous incident HD patient cohort, will be needed to confirm whether the impact of T3 levels on mortality is independent or associated with inflammation.

It is well known that T3 increases cardiac output by affecting tissue oxygen consumption, vascular resistance, blood volume, cardiac contractility, and heart rate (28, 29). Therefore, low T3, even in the normal range, has been suggested to be associated with various CVDs in ESRD patients, which has been clarified by several previous studies (12, 17, 30, 31). In their study, Zoccali et al. (12) showed that LV systolic function was depressed and LVMI was increased in the first fT3 quartile of 234 prevalent dialysis patients when compared with patients in other quartiles. In another recent, cross-sectional study, across decreasing fT3 tertiles, carotid-femoral pulse wave velocity (c-f PWV) and carotid artery-intima media thickness (CA-IMT) measured by Doppler ultrasonography were found to be incrementally higher in 137 maintenance HD patients. Furthermore, fT3 concentrations have been demonstrated to be significantly associated with c-f PWV and CA-IMT, indicating that fT3 levels are negatively associated with carotid atherosclerosis and arterial stiffness (30). The present study also found that in the ‘lower’ T3 group, LVEF was significantly lower, whereas LVMI was significantly higher, which was in concordance with previous observations. In addition, serum T3 concentrations had a significant inverse correlation with LVMI and a significant positive relationship with LVEF. Even though a multivariate Cox hazards regression analysis revealed that low T3 was a significant risk factor for mortality after adjusting for traditional risk factors, the independent predictive value of low T3 for mortality disappeared when LVEF and/or LVMI were included in a multivariate Cox model. These findings suggest that the association between T3 levels and all-cause mortality is largely mediated by cardiac dysfunction. Similarly, cardiac abnormalities, low LVEF and high LVMI had a significant impact on all-cause mortality only when T3 levels were excluded from the multivariate analysis (data not shown), suggesting that the relationship between cardiac dysfunction and mortality was mediated by T3.

A recent study by Ozen et al. (6) has shown that fT3 is a significant predictor of mortality in adjusted Cox regression models including albumin or hs-CRP levels, but that the significance is less when compared with that of a crude Cox model. Moreover, further adjustment for both albumin and hs-CRP levels made the impact of fT3 levels on mortality disappear. These findings suggest that fT3 is a strong and inverse mortality predictor, in part explained by its underlying association with nutritional state and inflammation. In this study, we included for the first time not only albumin levels but also the SGA score, LBM-Cr, and nPCR to determine the relationship between T3 and nutritional status and demonstrated that T3 levels were significantly associated with nutritional parameters in incident HD patients. Furthermore, T3 was a significant independent predictor of all-cause mortality in a traditional risk factors-adjusted Cox model. However, the relationship became insignificant when nutritional parameters were included in the multivariate Cox hazards regression analysis. Similarly, the impact of nutritional parameters on mortality was no longer significant when T3 was included in the Cox model (data not shown). Taken together, the prognostic power of T3 and nutrition seem to be interrelated.

Several aspects of the methods and scope of this study may limit the interpretation of research outcomes. First, since the study subjects were all Korean incident HD patients, the association between T3 levels and mortality may not be generalized to other populations. Second, we focused on studying the effect of T3 levels in this study, while previous investigations have found that both low T3 and fT3 levels are independent risk factors for mortality in prevalent dialysis patients (6, 17, 18, 32). In incident dialysis patients, however, only T3, but not fT3, levels were associated with worse all-cause and CV mortality after adjustment for confounding factors (31). This may be attributed to a lower technical specificity of fT3 measurement by RIA (20), especially for the stored samples (21), and a smaller influence of heparin on T3 than on fT3 (22, 33). In addition, T3 levels are determined as a total form rather than as a free form in most hospitals in Korea due to the financial benefits and convenience. Based on these findings, we considered T3 as a more stable and reliable factor, and used T3 rather than fT3 for the analysis. Third, more than half of our study population, mostly having CV complications, were taking beta blockers. Patients with renal failure, especially those on dialysis treatment, have a considerably high prevalence of CV complications due to uremic toxin, volume overload, chronic inflammation, oxidative stress, and abnormal bone metabolism (34). In fact, 87% of the patients had hypertension and 36.5% of the patients had underlying CV complications in this study. Even though beta blockers are known to affect serum T3 concentrations, we inferred that excluding these patients would result in selection bias. Moreover, the proportion of patients on beta blockers was not different between the ‘lower’ and ‘higher’ T3 groups, and beta blocker usage was not significantly associated with serum T3 concentrations in the correlation analysis (r=0.062, P=0.180). Furthermore, in the linear regression analysis, beta blocker usage was not a significant determinant of serum T3 concentrations (R=0.043, P=0.093). Fourth, there was a disparity in the timing of laboratory measurements and echocardiography. As has been mentioned above, since there was an extremely wide variation in the biochemical data at 0 month due to diverse patients' conditions to start HD, we used the laboratory data collected at 3 months after HD commencement, when correction of uremia, anemia, and acidosis was considered to be sufficiently achieved. On the other hand, based on the results of previous studies demonstrating that LVMI, LVEF, and fractional shortening are improved after long-term HD of 12–36 months (35, 36), correction of malnutrition, anemia, acidosis, and thyroid hormone might influence the cardiac function at 3 months. However, no studies have yet shown significant changes in echocardiographic findings within 3 months of short-term dialysis. In this study, echocardiography was carried out not at the start of dialysis but close to the time of discharge, when patients were considered to be in euvolemic state. In fact, the mean interval between dialysis commencement and echocardiography was 19.7 days in our patients. In addition, since most incident HD patients had to undergo maintenance HD at private clinics in Korea and to pay an additional fee of $300–350 for echocardiography, it was not easy to perform echocardiography at outpatient clinics after 3 months of initial dialysis. Therefore, we inevitably used 3-month laboratory data and baseline (about 3 weeks after the start of HD) echocardiographic parameters. Fifth, since only 3-month laboratory data were included in the analysis, it was difficult to determine the impact of the change in T3 levels on patients' clinical outcomes. Meuwese et al. (32) assessed the association between baseline and trimestral variation in T3 levels and mortality in 210 prevalent HD patients and found that not only basal low T3 but also persistently low T3 was significantly associated with increased mortality risk, especially due to CV causes. While the HRs for all-cause and CV mortality were higher in patients with persistently low T3 than in patients with low basal T3, there was still a significant association of low basal T3 with mortality. Sixth, all-cause and especially CV mortality rates in this study were relatively low. A small number of events may limit the power of statistical analyses in identifying the independent predictors of all-cause and especially CV mortality. Finally, the follow-up duration was relatively short and thus a sufficient number of patients could not be recruited for a longitudinal study. Since this prospective observation cohort study is still ongoing, the long-term association between T3 levels and mortality can be elucidated in the future. Despite these limitations, to our knowledge, this study is the first to investigate the association of T3 levels with echocardiographic findings and various nutritional parameters, including albumin levels, SGA score, LBM-Cr, and nPCR, and their impact on all-cause and CV mortality in a large, ethnically homogenous incident HD patient cohort.

In conclusion, low T3 is a significant independent risk factor for all-cause mortality in incident HD patients, and its impact on mortality may be attributed, in part, to malnutrition and cardiac dysfunction.

Supplementary data

This is linked to the online version of the paper at http://dx.doi.org/10.1530/EJE-13-0540.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.

Acknowledgements

The authors acknowledge the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (A084001), for the support provided through a grant.

References

  • 1

    Iglesias P, Diez JJ. Thyroid dysfunction and kidney disease. European Journal of Endocrinology 2009 160 503515. (doi:10.1530/EJE-08-0837).

  • 2

    den Hollander JG, Wulkan RW, Mantel MJ, Berghout A. Correlation between severity of thyroid dysfunction and renal function. Clinical Endocrinology 2005 62 423427. (doi:10.1111/j.1365-2265.2005.02236.x).

    • Search Google Scholar
    • Export Citation
  • 3

    Shin DH, Lee MJ, Kim SJ, Oh HJ, Kim HR, Han JH, Koo HM, Doh FM, Park JT, Han SH. Preservation of renal function by thyroid hormone replacement therapy in chronic kidney disease patients with subclinical hypothyroidism. Journal of Clinical Endocrinology and Metabolism 2012 97 27322740. (doi:10.1210/jc.2012-1663).

    • Search Google Scholar
    • Export Citation
  • 4

    van Hoek I, Daminet S. Interactions between thyroid and kidney function in pathological conditions of these organ systems: a review. General and Comparative Endocrinology 2009 160 205215. (doi:10.1016/j.ygcen.2008.12.008).

    • Search Google Scholar
    • Export Citation
  • 5

    Song SH, Kwak IS, Lee DW, Kang YH, Seong EY, Park JS. The prevalence of low triiodothyronine according to the stage of chronic kidney disease in subjects with a normal thyroid-stimulating hormone. Nephrology, Dialysis, Transplantation 2009 24 15341538. (doi:10.1093/ndt/gfn682).

    • Search Google Scholar
    • Export Citation
  • 6

    Ozen KP, Asci G, Gungor O, Carrero JJ, Kircelli F, Tatar E, Sevinc Ok E, Ozkahya M, Toz H, Cirit M. Nutritional state alters the association between free triiodothyronine levels and mortality in hemodialysis patients. American Journal of Nephrology 2011 33 305312. (doi:10.1159/000324883).

    • Search Google Scholar
    • Export Citation
  • 7

    Chopra IJ. Nonthyroidal illness syndrome or euthyroid sick syndrome? Endocrine Practice 1996 2 4552. (doi:10.4158/EP.2.1.45).

  • 8

    Wiederkehr MR, Kalogiros J, Krapf R. Correction of metabolic acidosis improves thyroid and growth hormone axes in haemodialysis patients. Nephrology, Dialysis, Transplantation 2004 19 11901197. (doi:10.1093/ndt/gfh096).

    • Search Google Scholar
    • Export Citation
  • 9

    Zoccali C, Tripepi G, Cutrupi S, Pizzini P, Mallamaci F. Low triiodothyronine: a new facet of inflammation in end-stage renal disease. Journal of the American Society of Nephrology 2005 16 27892795. (doi:10.1681/ASN.2005040356).

    • Search Google Scholar
    • Export Citation
  • 10

    Malyszko J, Malyszko JS, Pawlak K, Mysliwiec M. Thyroid function, endothelium, and inflammation in hemodialyzed patients: possible relations? Journal of Renal Nutrition 2007 17 3037. (doi:10.1053/j.jrn.2006.07.003).

    • Search Google Scholar
    • Export Citation
  • 11

    Yilmaz MI, Sonmez A, Karaman M, Ay SA, Saglam M, Yaman H, Kilic S, Eyileten T, Caglar K, Oguz Y. Low triiodothyronine alters flow-mediated vasodilatation in advanced nondiabetic kidney disease. American Journal of Nephrology 2011 33 2532. (doi:10.1159/000322581).

    • Search Google Scholar
    • Export Citation
  • 12

    Zoccali C, Benedetto F, Mallamaci F, Tripepi G, Cutrupi S, Pizzini P, Malatino LS, Bonanno G, Seminara G. Low triiodothyronine and cardiomyopathy in patients with end-stage renal disease. Journal of Hypertension 2006 24 20392046. (doi:10.1097/01.hjh.0000244954.62362.8f).

    • Search Google Scholar
    • Export Citation
  • 13

    Kozdag G, Ural D, Vural A, Agacdiken A, Kahraman G, Sahin T, Ural E, Komsuoglu B. Relation between free triiodothyronine/free thyroxine ratio, echocardiographic parameters and mortality in dilated cardiomyopathy. European Journal of Heart Failure 2005 7 113118. (doi:10.1016/j.ejheart.2004.04.016).

    • Search Google Scholar
    • Export Citation
  • 14

    Pingitore A, Landi P, Taddei MC, Ripoli A, L'Abbate A, Iervasi G. Triiodothyronine levels for risk stratification of patients with chronic heart failure. American Journal of Medicine 2005 118 132136. (doi:10.1016/j.amjmed.2004.07.052).

    • Search Google Scholar
    • Export Citation
  • 15

    Scoscia E, Baglioni S, Eslami A, Iervasi G, Monti S, Todisco T. Low triiodothyronine (T3) state: a predictor of outcome in respiratory failure? Results of a clinical pilot study. European Journal of Endocrinology 2004 151 557560. (doi:10.1530/eje.0.1510557).

    • Search Google Scholar
    • Export Citation
  • 16

    Malik R, Hodgson H. The relationship between the thyroid gland and the liver. QJM: Monthly Journal of the Association of Physicians 2002 95 559569. (doi:10.1093/qjmed/95.9.559).

    • Search Google Scholar
    • Export Citation
  • 17

    Zoccali C, Mallamaci F, Tripepi G, Cutrupi S, Pizzini P. Low triiodothyronine and survival in end-stage renal disease. Kidney International 2006 70 523528. (doi:10.1038/sj.ki.5001566).

    • Search Google Scholar
    • Export Citation
  • 18

    Enia G, Panuccio V, Cutrupi S, Pizzini P, Tripepi G, Mallamaci F, Zoccali C. Subclinical hypothyroidism is linked to micro-inflammation and predicts death in continuous ambulatory peritoneal dialysis. Nephrology, Dialysis, Transplantation 2007 22 538544. (doi:10.1093/ndt/gfl605).

    • Search Google Scholar
    • Export Citation
  • 19

    Rotondi M, Netti GS, Rosati A, Mazzinghi B, Magri F, Ronconi E, Becherucci F, Pradella F, Salvadori M, Serio M. Pretransplant serum FT3 levels in kidney graft recipients are useful for identifying patients with higher risk for graft failure. Clinical Endocrinology 2008 68 220225. (doi:10.1111/j.1365-2265.2007.03022.x).

    • Search Google Scholar
    • Export Citation
  • 20

    Squire CR. Methods for the investigation of thyroid function. Methods in Molecular Biology 2006 324 91108. (doi:10.1385/1-59259-986-9-91).

    • Search Google Scholar
    • Export Citation
  • 21

    Kilinc AS, Duzoylum A, Uncugil CF, Yucel D. Falsely increased free triiodothyronine in sera stored in serum separator tubes. Clinical Chemistry 2002 48 22962297.

    • Search Google Scholar
    • Export Citation
  • 22

    Stevenson HP, Archbold GP, Johnston P, Young IS, Sheridan B. Misleading serum free thyroxine results during low molecular weight heparin treatment. Clinical Chemistry 1998 44 10021007.

    • Search Google Scholar
    • Export Citation
  • 23

    Michels WM, Grootendorst DC, Verduijn M, Elliott EG, Dekker FW, Krediet RT. Performance of the Cockcroft-Gault, MDRD, and new CKD-EPI formulas in relation to GFR, age, and body size. Clinical Journal of the American Society of Nephrology 2010 5 10031009. (doi:10.2215/CJN.06870909).

    • Search Google Scholar
    • Export Citation
  • 24

    Steiber AL, Kalantar-Zadeh K, Secker D, McCarthy M, Sehgal A, McCann L. Subjective Global Assessment in chronic kidney disease: a review. Journal of Renal Nutrition 2004 14 191200.

    • Search Google Scholar
    • Export Citation
  • 25

    Noori N, Kovesdy CP, Bross R, Lee M, Oreopoulos A, Benner D, Mehrotra R, Kopple JD, Kalantar-Zadeh K. Novel equations to estimate lean body mass in maintenance hemodialysis patients. American Journal of Kidney Diseases 2011 57 130139. (doi:10.1053/j.ajkd.2010.10.003).

    • Search Google Scholar
    • Export Citation
  • 26

    Jindal KK, Goldstein MB. Urea kinetic modelling in chronic hemodialysis: benefits, problems, and practical solutions. Seminars in Dialysis 1988 1 8285. (doi:10.1111/j.1525-139X.1988.tb00710.x).

    • Search Google Scholar
    • Export Citation
  • 27

    Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. American Journal of Cardiology 1986 57 450458. (doi:10.1016/0002-9149(86)90771-X).

    • Search Google Scholar
    • Export Citation
  • 28

    Klein I, Ojamaa K. Thyroid hormone and the cardiovascular system. New England Journal of Medicine 2001 344 501509. (doi:10.1056/NEJM200105103441901).

    • Search Google Scholar
    • Export Citation
  • 29

    Klein I, Danzi S. Thyroid disease and the heart. Circulation 2007 116 17251735. (doi:10.1161/CIRCULATIONAHA.106.678326).

  • 30

    Tatar E, Kircelli F, Asci G, Carrero JJ, Gungor O, Demirci MS, Ozbek SS, Ceylan N, Ozkahya M, Toz H. Associations of triiodothyronine levels with carotid atherosclerosis and arterial stiffness in hemodialysis patients. Clinical Journal of the American Society of Nephrology 2011 6 22402246. (doi:10.2215/CJN.02540311).

    • Search Google Scholar
    • Export Citation
  • 31

    Carrero JJ, Qureshi AR, Axelsson J, Yilmaz MI, Rehnmark S, Witt MR, Barany P, Heimburger O, Suliman ME, Alvestrand A. Clinical and biochemical implications of low thyroid hormone levels (total and free forms) in euthyroid patients with chronic kidney disease. Journal of Internal Medicine 2007 262 690701. (doi:10.1111/j.1365-2796.2007.01865.x).

    • Search Google Scholar
    • Export Citation
  • 32

    Meuwese CL, Dekker FW, Lindholm B, Qureshi AR, Heimburger O, Barany P, Stenvinkel P, Carrero JJ. Baseline levels and trimestral variation of triiodothyronine and thyroxine and their association with mortality in maintenance hemodialysis patients. Clinical Journal of the American Society of Nephrology 2012 7 131138. (doi:10.2215/CJN.05250511).

    • Search Google Scholar
    • Export Citation
  • 33

    Jaume JC, Mendel CM, Frost PH, Greenspan FS, Laughton CW. Extremely low doses of heparin release lipase activity into the plasma and can thereby cause artifactual elevations in the serum-free thyroxine concentration as measured by equilibrium dialysis. Thyroid 1996 6 7983. (doi:10.1089/thy.1996.6.79).

    • Search Google Scholar
    • Export Citation
  • 34

    Schiffrin EL, Lipman ML, Mann JF. Chronic kidney disease: effects on the cardiovascular system. Circulation 2007 116 8597. (doi:10.1161/CIRCULATIONAHA.106.678342).

    • Search Google Scholar
    • Export Citation
  • 35

    Covic A, Mardare NG, Ardeleanu S, Prisada O, Gusbeth-Tatomir P, Goldsmith DJ. Serial echocardiographic changes in patients on hemodialysis: an evaluation of guideline implementation. Journal of Nephrology 2006 19 783793.

    • Search Google Scholar
    • Export Citation
  • 36

    Foley RN, Parfrey PS, Kent GM, Harnett JD, Murray DC, Barre PE. Serial change in echocardiographic parameters and cardiac failure in end-stage renal disease. Journal of the American Society of Nephrology 2000 11 912916.

    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

 

     European Society of Endocrinology

Sept 2018 onwards Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 503 56 4
PDF Downloads 236 85 8
  • View in gallery

    The correlation of serum T3 levels with various nutritional, inflammatory, and echocardiographic variables. Serum T3 concentrations had significant positive associations with (a) serum albumin levels, (b) LBM-Cr, (c) nPCR, and (d) LVEF, while there were significant inverse correlations of T3 with (e) LVMI and (f) Ln hs-CRP. LBM-Cr, lean body mass estimated by creatinine kinetics; nPCR, normalized protein catabolic rate; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; Ln hs-CRP, log-transformed high-sensitivity C-reactive protein.

  • View in gallery

    Kaplan–Meier plots for all-cause and cardiovascular survival. Compared with patients with ‘higher’ T3, the ‘lower’ T3 group showed significantly worse (a) all-cause and (b) cardiovascular cumulative survival rates.

  • View in gallery

    Receiver operating characteristic curves of T3 for all-cause mortality. Serum T3 concentrations provided significant positive predictive values for all-cause mortality (a) in an unadjusted model (AUC=0.674) and (b) in a model adjusted for BMI, smoking status, modified CCI, aspirin and clopidogrel usage, serum HbA1c, albumin, and potassium concentrations, eGFR by CKD-EPI, Ln hs-CRP, and SGA score (AUC=0.878). CCI, Charlson comorbidity index; eGFR, estimated glomerular filtration rate; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; Ln hs-CRP, log-transformed high-sensitivity C-reactive protein; SGA, subjective global assessment.