Author + information
- Received February 28, 1997
- Revision received June 26, 1997
- Accepted July 1, 1997
- Published online October 1, 1997.
- Stefan D Anker, MDAB,1,
- Andrew L Clark, MD, MRCPA,*,
- Michael Kemp, MSc, MRCPathC,
- Colin Salsbury, MSc, FIBMSC,
- Mauro M Teixeira, PhDA,
- Paul G Hellewell, PhDC and
- Andrew J.S Coats, DM, FRCP, FACCA
- ↵*Dr. Andrew L. Clark, Department of Cardiology, Western Infirmary, Dumbarton Road, Glasgow G11 6NT, Scotland, United Kingdom.
Objectives. We sought to assess the possible relations between clinical severity of chronic heart failure and catabolic factors, specifically tumor necrosis factor (TNF), soluble TNF receptors 1 and 2 (sTNFR-1 and sTNFR-2), cortisol, testosterone and dehydroepiandrosterone (DHEA).
Background. Chronic heart failure is associated with loss of muscle bulk that may be related to alteration of the balance between catabolism and anabolism.
Methods. Sixty-three patients (average age ±SD 60.4 ± 11.3 years) with stable chronic heart failure and 20 control subjects aged 52.8 ± 11.4 years were studied. We measured body mass index (BMI) and obtained maximal incremental exercise testing with metabolic gas exchange measurements and measurements of venous levels of TNF, sTNFR-1 and sTNFR-2, cortisol and DHEA.
Results. There was no difference in total TNF-alpha levels between patients and control subjects (9.76 ± 8.59 vs. 6.84 ± 2.7 pg/ml). sTNFR-1 (128.9 ± 84.5 vs. 63.6 ± 23.3 pg/ml, p < 0.003) and sTNFR-2 (250.1 ± 109.5 vs. 187.9 ± 92.2 pg/ml, p = 0.03) were higher in patients. DHEA was lower in patients (9.88 ± 6.94 vs. 15.64 ± 8.33 nmol/liter, p = 0.004). The ratio of log cortisol to log DHEA correlated with log TNF level (r = 0.50, p < 0.001 for the patients alone; r = 0.48, p < 0.001 for the group as a whole). Peak oxygen consumption correlated with both sTNFR-1 and sTNFR-2 (r = −0.51, p < 0.001 and r = −0.39, p < 0.001, respectively). There was a negative correlation between BMI and TNF levels (r = −0.43, p < 0.001 for the patients) and the cortisol/DHEA ratio (r = −0.32, p = 0.01 for the patients).
Conclusions. There is an increase in TNF and its soluble receptors in chronic heart failure. This increase is associated with a rise in the cortisol/DHEA (catabolic/anabolic) ratio. These changes correlate with BMI and clinical severity of heart failure, suggesting a possible etiologic link.
The origin of the weight loss and cachexia seen in patients with chronic heart failure is still unclear. Evidence of muscle wasting is present even in mild chronic heart failure , and the muscle wasting appears to be related to exercise capacity . Cachectic patients with chronic heart failure are weaker than noncachectic patients despite having similar conventional measures of disease severity . Possible mediators of weight loss include inflammatory cytokines, and interest has focused on the possible role of tumor necrosis factor-alpha (TNF-alpha) [4, 5]. TNF may have further significance as a myocardial depressant [6, 7]. TNF-alpha appears to mediate its biologic effects by way of two different cell membrane receptors . The TNF-binding portion of the receptors can be found in solution, known as sTNFR-1 and sTNFR-2. The soluble receptors are increased after acute myocardial infarction . Some investigators have found an increase in the sTNFRs in chronic heart failure with a reduction in cardiac myocyte–bound receptors.
Other possible mediators of weight loss include alterations in adrenal hormone metabolism. In tuberculosis, wasting is associated with an increase in the ratio between cortisol and the anabolic hormone, dehydroepiandrosterone (DHEA) . Whether this change in steroid metabolism is caused by TNF activation is not clear , but it may be related to patterns of immune activation .
We set out to examine the possible relations between TNF, its receptors and the generalized neurohormonal activation of chronic heart failure. We also investigated the possible role of altered steroid metabolism and, in particular, the cortisol/DHEA ratio as a measure of the endocrine balance between anabolism and catabolism. Testosterone, another anabolic steroid, was also measured. We also attempted to relate any changes we found to the clinical state and exercise performance of the patients.
The study was approved by the ethics committee of the Royal Brompton National Heart and Lung Hospital. All subjects gave fully informed signed consent.
We studied 63 patients (average age ± SD 60.4 ± 11.3 years) and 20 control subjects aged 52.8 ± 11.4 years. Of the patients, 26 had dilated cardiomyopathy and the other 37 had ischemic cardiomyopathy. Baseline demographic data are shown in Table 1. All patients had been clinically stable for a minimum of 3 months before study. No patient was edematous and all were on a medication regimen that had not changed for ≥2 months. No patient had either infection or neoplastic disease at the time of investigation. The average dose of diuretic agent used was 115.6 ± 95.9 mg daily (where 1 mg of bumetanide was considered equivalent to 40 mg of furosemide). Twelve patients were not taking an angiotensin-converting enzyme inhibitor. Of the others, 24 were receiving captopril, 21 enalapril, 5 lisinopril and 1 ramipril. Twenty-three patients were receiving digoxin, 17 amiodarone, 22 warfarin and 24 aspirin. All patients were limited by either breathlessness or fatigue, or both.
Control subjects were recruited from hospital staff members and subjects presenting for a routine health examination. Subjects were only used as control subjects if they had no past medical history of cardiac or pulmonary or neoplastic disease and if a rest electrocardiogram and pulmonary function tests were normal.
Height and nude weight were measured in all subjects. Subjects underwent maximal incremental exercise testing using a standard Bruce protocol with the addition of a stage 0, that is, 3 min at 1.7 km/h at zero gradient. Patients breathed through a one-way valve connected to a respiratory mass spectrometer (Innovision, Odense, Denmark). Ventilation (V̇e), oxygen consumption (V̇o2) and carbon dioxide production (V̇co2) were measured on-line every 10 s. Exercise was started after basal readings had been stable for ≥2 min. Subjects were encouraged to exercise to exhaustion. Peak V̇o2was taken as an index of exercise capacity. The slope of the relation between ventilation and carbon dioxide production (V̇e/V̇co2slope) was calculated from the exercise data and taken as an index of the ventilatory response to exercise.
Blood samples were taken for hormone and cytokine assay before the exercise test. Blood was withdrawn between 9 and 10 amafter a 12-h fast in a quiet, darkened room. A polyethylene cannula was inserted into an antecubital vein, and blood was withdrawn after the subject had been lying supine for 20 min. Blood was centrifuged immediately and serum stored at −70°C for later analysis. Blood was also analyzed for serum creatinine, urea, albumin and total protein as indexes of nutritional state.
Radioimmunoassays were performed for renin (Biodata, Gudonia Montecelio, Italy) and aldosterone (DPC). Catecholamines were measured by high performance liquid chromatography. TNF was measured by an ELISA assay with a lower limit of detectability of 3.0 pg/ml (Medgenix, Fleurus, Belgium) which measured total TNF, both free and bound to soluble receptors. sTNFR-1 (lower limit of detection 25 pg/ml) and sTNFR-2 (lower limit of detection 2 pg/ml) were measured by using test kits from R&D Systems. Cortisol was measured using an ELISA assay (Boeringer Mannheim, Germany) and DHEA was measured by radioimmunoassay with a lower limit of detectability of 0.04 ng/ml (DPC).
1.1 Statistical methods.
Comparisons between patients and control subjects were made by Student ttest. The cytokine measurements were log-normally distributed, and comparisons were made between sets of log-transformed data. Linear regression using the least squares method was used to establish correlations between variables.
All subjects managed to exercise to a point where the respiratory exchange ratio (V̇co2/V̇o2) exceeded 1.0, indicating that at least near maximal exercise capacity was reached. Patients had a lower peak V̇o2and higher V̇e/V̇co2slope than control subjects. There was no significant difference in the V̇co2/V̇o2ratio achieved at peak exercise.
There was evidence of general activation of the sympathetic and renin-angiotensin systems in patients with chronic heart failure (Table 2) with higher levels of aldosterone (729.4 ± 695.3 vs. 283.5 ± 167.2 pmol/liter, p = 0.006), renin (12.06 ± 13.27 vs. 1.36 ± 0.71 ng/ml per h, p = 0.001), norepinephrine (3.67 ± 2.5 vs. 1.77 ± 0.72 nmol/liter, p = 0.001) and epinephrine (1.38 ± 1.68 vs. 0.50 ± 0.16 mmol/liter, p = 0.022) than in control subjects. Urea and creatinine were higher in the patients (Table 3). Albumin was lower and total protein higher in the patients.
Levels of TNF were not significantly higher in the patients, but levels of both TNF receptors were higher in patients than in control subjects. There were no differences in cytokine levels between patients with ischemic versus idiopathic cardiomyopathy.
Within the patient group, levels of the two TNF receptors correlated (r = 0.80, p < 0.001) and each correlated with the level of TNF (sTNFR-1, r = 0.57, p < 0.001; sTNFR-2, r = 0.52, p < 0.001). These correlations were unchanged when recalculated for the group as a whole (r = 0.83, r = 0.50 and r = 0.44, respectively, p < 0.001 for the three comparisons). The log[cortisol]/log[DHEA] ratio was higher in patients than in control subjects, but there was no difference between the two groups in testosterone and estradiol levels. The ratio of log [cortisol]/log[DHEA] correlated with log[TNF] level (r = 0.50, p < 0.001 for the patients alone; r = 0.48, p < 0.001 for the group as a whole) (Fig. 1).
2.1 Relations with clinical variables.
There was no relation between left ventricular ejection fraction and any of the cytokines measured. There was a relation between peak V̇o2and both sTNFR-1 and sTNFR-2 (r = −0.54, p < 0.001 and r = −0.39, p < 0.001, respectively) (Fig. 2). There were similar relations between V̇e/V̇co2slope and both sTNFR-1 and sTNFR-2 (r = 0.49, p < 0.001 and r = 0.39, p < 0.001, respectively). Although TNF was not significantly higher in the patient group, it correlated with peak V̇o2(r = −0.34, p < 0.002) and weakly with New York Heart Association (NYHA) functional class (r = 0.22, p < 0.05). There was a weak relation between testosterone and DHEA (r = −0.33, p < 0.001) but not between testosterone and cortisol.
There was a negative correlation between body mass index (BMI) and TNF levels (r = −0.43, p < 0.001 for the patients alone; r = −0.40, p < 0.001 for the group as a whole). There was a weak relation between BMI and the cortisol/DHEA ratio (r = −0.32, p = 0.01 for the patients alone; r = −0.25, p = 0.03 for the group as a whole).
There was a relation between serum creatinine and sTNFR-1 (r = 0.70, p < 0.001), sTNFR-2 (r = 0.60, p < 0.001) and TNF (0.43, p < 0.001) (Fig. 3). These relations were much stronger than those between serum albumin and sTNFR-1 (r = −0.39, p < 0.001), sTNFR-2 (r = −0.27, p = 0.02) and TNF (−0.27, p = 0.02).
The weight loss of patients with chronic heart failure has been recognized since earliest times , but the cause remains obscure. A variety of abnormalities of cytokines have been described [4, 5]as has an increase in insulin resistance and an elevation in catabolic steroids in untreated patients . This alteration in body metabolism suggests a possible mechanism by which weight loss may occur. We wished to assess the catabolic state in a group of patients with chronic heart failure by measuring TNF and its receptors, and the ratio of catabolic to anabolic steroid hormones.
In common with previous investigators , although the TNF level was not significantly raised in the patient group, we found an increasing level of TNF with worsening severity of chronic heart failure as assessed by peak V̇o2and NYHA symptom class, although this association is weak. There were stronger relations between clinical severity of heart failure and levels of both TNF receptors. The level of TNF correlated most closely with BMI. The ratio of catabolic to anabolic steroid as measured by cortisol/DHEA ratio also correlated with body mass.
The cause of the rise in catabolic factors is unclear. We found that levels of TNF receptors correlated with serum creatinine, suggesting a possible role for decreased excretion. The role of soluble TNF receptors is not clear. They are increased in chronic heart failure, whereas myocyte receptors appear to be decreased . It may be that the elevated sTNF receptor levels represent a reservoir of TNF in the circulation, prolonging its biologic effects. The mechanisms by which changes in cytokines and hormones might lead to loss of muscle bulk are also unclear, but cytokine-induced programmed cell death may be occurring in skeletal muscle.
The patterns of steroid metabolism have not previously been described in chronic heart failure. Alteration in the cortisol/DHEA ratio may have important effects on the patterns of immune activation , influencing the balance of T helper cell maturation between T helper cell 1 and T helper cell 2 pathways. The alteration in balance between anabolic and catabolic steroids may have consequences for muscle metabolism and wasting. We found the cortisol/DHEA ratio correlated with body mass, and in tuberculosis, this relation also appears to be related to loss of muscle bulk . TNF release from macrophages appears to be affected by relatively small changes in sex hormone concentration , but we did not find a consistent pattern of alteration in sex hormone levels in our group of patients.
3.1 Limitations to present study.
This is an observational study and cannot prove causal relations between levels of cytokines and hormones and the progression of the heart failure syndrome and loss of muscle bulk. It may be that increased levels of TNF and sTNFRs are merely reflections of severity of illness. Given the present knowledge of the biologic effects of these factors, it seems reasonable to suggest an etiologic role in the muscle wasting associated with chronic heart failure.
We have measured total, rather than free, testosterone, and we may therefore have missed any relation between testosterone and other variables if significant variations in sex steroid binding globulin were present. Our approach is inevitably limited by the intrinsic circadian variability in some of the substances measured and their sensitivity to environmental changes. Little is known regarding the diurnal variation of cytokine and steroid levels in chronic heart failure. For that reason, we ensured that blood samples were drawn under standard conditions at the same time of day in all subjects. Further, levels of TNF, although elevated in chronic heart failure, fluctuate markedly. In a study of 16 patients followed up for >1 year , TNF was undetectable in every patient on at least one occasion, although the sensitivity of the test kit used was lower than that used in our study.
We found an increase in TNF and soluble TNF receptors that correlates with severity of heart failure and with body mass. In addition, we have shown an alteration in the cortisol/DHEA ratio, suggesting a preferential effect of catabolism over anabolism. These changes may be causally related to weight loss and muscle wasting in chronic heart failure.
↵1 Dr. Anker is a Research Fellow of the European Society of Cardiology, Rotterdam, The Netherlands.
☆ Dr. Clark is supported in part by a fellowship from the Robert Luff Foundation and Dr. Coats by the Viscount Royston Trust, London, England, United Kingdom.
- body mass index
- New York Heart Association
- soluble tumor necrosis factor receptor
- tumor necrosis factor
- carbon dioxide production
- oxygen consumption
- Received February 28, 1997.
- Revision received June 26, 1997.
- Accepted July 1, 1997.
- The American College of Cardiology
- Mancini DM,
- Walter G,
- Reichek N,
- et al.
- Volterrani M,
- Clark AL,
- Ludman PF,
- et al.
- Anker SD,
- Swan JW,
- Volterrani M,
- et al.
- McMurray J,
- Abdullah I,
- Dargie HJ,
- Shapiro D
- Finkel MS,
- Oddis CV,
- Jacob TD,
- et al.
- Torre-Amione G,
- Kapadia S,
- Lee J,
- et al.
- Rook GAW,
- Honour J,
- Kon OM,
- Wilkinson RJ,
- Davidson R,
- Shaw RJ
- Grange JM
- Daynes RA,
- Araneo BA,
- Hennebold J,
- Enioutina E,
- Hong HM
- Katz AM,
- Katz PB
- Swan JW,
- Walton C,
- Godsland IF,
- Clark AL,
- Coats AJS
- Anand IA,
- Ferrari R,
- Kalra GS,
- Wahi PL,
- Poole-Wilson PA,
- Harris PC
- Torre-Amione G,
- Kapadia S,
- Benedict C,
- Oral H,
- Young JB,
- Mann DL
- Dutka DP,
- Elborn JS,
- Delamere F,
- Shale DJ,
- Morris GK