SEARCH

SEARCH BY CITATION

Keywords:

  • chronic fatigue syndrome;
  • heat shock proteins;
  • maximal exercise;
  • oxidative stress;
  • severe infection;
  • sport practice

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Abstract.  Jammes Y, Steinberg JG, Delliaux S (Aix-Marseille University, Marseille, France). Chronic fatigue syndrome: acute infection and history of physical activity affect resting levels and response to exercise of plasma oxidant/antioxidant status and heat shock proteins. J Intern Med 2012; 272: 74–84.

Objectives.  A history of high-level physical activity and/or acute infection might constitute stress factors affecting the plasma oxidant–antioxidant status and levels of heat shock proteins (HSPs) in patients with chronic fatigue syndrome (CFS).

Design.  This case–control study compared data from 43 CFS patients to results from a matched control group of 23 healthy sedentary subjects.

Setting and subjects.  Five patients had no relevant previous history (group I). Eighteen had practised high-level sport (group II), and severe acute infection had been diagnosed in nine patients (group III). A combination of sport practice and infection was noted in 11 patients (group IV).

Interventions.  After examination at rest, all subjects performed a maximal cycling exercise test. Plasma levels of two markers of oxidative stress [thiobarbituric acid reactive substances (TBARS) and reduced ascorbic acid (RAA)] and both HSP27 and HSP70 were measured.

Results.  At rest, compared with the control group, the TBARS level was higher in groups II, III and IV patients, and the RAA level was lower in groups III and IV. In addition, HSP70 levels were significantly lower in all CFS groups, compared with controls, but negative correlations were found between resting HSP27 and HSP70 levels and the history of physical activity. After exercise, the peak level of TBARS significantly increased in groups II, III and IV, and the variations in HSP27 and HSP70 were attenuated or suppressed, with the greatest effects in groups III and IV.

Conclusion.  The presence of stress factors in the history of CFS patients is associated with severe oxidative stress and the suppression of protective HSP27 and HSP70 responses to exercise.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

In patients with chronic fatigue syndrome (CFS), we have previously shown a marked increase in exercise-induced oxidative stress with a reduced heat shock protein (HSP) response [1]. This corroborates the observation of an acute relationship between the redox status and HSP expression [2], suggesting that the reduction in HSP expression increases the generation of reactive oxygen species (ROS). We previously hypothesized that the accumulation of two or more stressors might be the cause of a lower HSP production in later life [1].

Chronic fatigue syndrome results in a severe impairment in sport and daily activities [3–5], and exercise-induced aetiopathology is suspected [6, 7]. Indeed, CFS is sometimes diagnosed in elite cyclists [8] and in later life in subjects who had previously exercised frequently [9]. CFS also often follows severe bacterial or viral infection [10–12] with various pathogens, including Epstein–Barr virus, cytomegalovirus, human herpes virus, enterovirus, parvovirus and mycoplasma [10]. Because both physical activity [13, 14] and infection [15, 16] are responsible for an oxidative stress and HSP response in healthy subjects, it was hypothesized that these stressors might have a role in the aetiopathogenesis of CFS [4].

Studies in patients with CFS indicate an excessive production of ROS following exertion [1, 17–19] and changes in resting blood oxidant–antioxidant status [20–23]. Vecchiet et al. [23] reported that the muscle symptoms of fatigue were proportional to the blood levels of a marker of ROS-induced lipid peroxidation [thiobarbituric acid reactive substances (TBARS)]. Excessive ROS production during and after muscle contraction at a high strength exerts well-known effects, including inhibition of Na+–K+-pump activity and membrane lipid peroxidation [24] and activation of the chemosensitive group IV muscle afferents [25]. This group detects almost all the metabolites produced by the contracting muscle (including K+, lactic acid, myokines and ROS) and contributes to pain sensation and the sensorimotor control of muscles. Several studies have demonstrated a correlation between musculoskeletal symptoms and an accentuated lipid peroxidation at rest in CFS patients [23, 26, 27].

In healthy subjects, excessive ROS production may also elicit local and systemic inflammatory responses with enhanced production of cytokines [28]. The elevated ROS production in CFS patients is sometimes associated with altered innate immunity and an accentuated inflammatory response to exercise [29–36]. However, no clear evidence of a link between abnormal immunity and CFS has been established [1, 10].

In the present study, we have evaluated in CFS patients the history of sport practice and/or episodes of severe acute infection in relation to the resting plasma oxidant–antioxidant status and HSP27 and HSP70 levels, and the maximal exercise-induced oxidative stress and HSP response. The resting level of interleukin (IL)-1β was also measured to highlight any association between inflammation and oxidative stress. The history of sport practice was self-reported by patients, whereas the diagnosis of infection was made and reported by their physicians. Our objective was to identify which of these two stress factors coincides with the most significant changes in resting levels and/or exercise-induced changes in oxidant–antioxidant status and HSP response.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Subjects

Forty-three CFS patients participated in this study. To be diagnosed with CFS [3–5], patients had to have persistent or relapsing fatigue for at least six consecutive months, and four or more of the following eight symptoms: postexertional malaise, impaired memory or concentration, unrefreshing sleep, muscle pain, multijoint pain, tender cervical or axillary lymph nodes, sore throat and headache. Retrospectively, four groups of patients were identified with regard to their history of sport practice and/or severe acute infection. No relevant history could be identified in four patients (group I: no identifiable cause of CFS). Eighteen patients had practised sport at a high level (>6 h per week) for more than 6 years prior to symptom onset (group II: physical hyperactivity). In nine patients who had not practised any sport, severe infection (peritonitis, pneumonia or encephalomyelitis) had occurred within 3–4 months preceding the onset of CFS symptoms (group III: infection). In 11 patients who exercised frequently, an episode of severe infection had also been diagnosed (group IV: hyperactivity plus infection). CFS patients in groups II and IV had stopped or markedly reduced their sport practice 3 ± 2 years before they were examined in our laboratory.

The morphological characteristics and clinical data of CFS patients are shown in Table 1.

Table 1.   Characteristics of patients in chronic fatigue syndrome groups
 No history Group IPhysical hyperactivity Group IIInfection Group IIIHyperactivity plus infection Group IV
  1. ARDS, acute respiratory distress syndrome.

  2. Values are the mean ± SEM.

n518911
Sex ratio (female/male)5/08/104/56/5
Age (years)39 ± 838 ± 344 ± 638 ± 4
Weight (kg)62 ± 272 ± 369 ± 365 ± 4
Height (cm)162 ± 23163 ± 38177 ± 28164 ± 10
Duration of symptoms (years)6 ± 13 ± 0.53 ± 0.55 ± 2
Myalgia (n)2736
Sleep disorders1536
Yearly physical activity (h per year)<100439 ± 57<100448 ± 66
Infection (n)002 peritonitis 1 pneumonia 2 ARDS 4 encephalomyelitis1 peritonitis 4 pneumonia 3 ARDS 3 encephalomyelitis

Data from the CFS patients were compared with those obtained in a control group of 23 healthy Caucasian volunteers matched for gender, age, weight and socio-economic status (nine women and 14 men; mean age 39 ± 5 years; mean weight 74 ± 3 kg) who visited the clinic for a medical check-up. The control subjects were considered to be sedentary because they did not participate in regular formal exercise, only playing tennis or jogging for a maximum of 2 h per week. Procedures were carried out with the written informed consent of the subjects. Additional biochemical analyses of venous blood samples were approved by the ethics committee of our institution, and the protocol was performed in accordance with the Declaration of Helsinki.

Study design

All patients were referred to our laboratory by clinicians who had identified the symptoms of and diagnosed severe infection. On the same day, subjects were first questioned about their history of sport practice and underwent examination at rest, including electrocardiogram (ECG) recordings, arterial blood pressure measurements using a sphygmomanometer, and insertion of a catheter (Neofly 21G; Viggo-Spectramed, Miami, FL, USA) in an antecubital vein for measurements of blood concentrations of K+, lactic acid, TBARS, reduced ascorbic acid (RAA), IL-1β, HSP27 and HSP70. Recent data from routine biochemical analyses (glucose, cholesterol and electrolytes) were reported by physicians. The baseline levels of all biochemical variables are reported in Table 2. Next, patients performed a maximal incremental cycling exercise test, as in our previous studies [1, 13, 17], with measurements of cardiorespiratory variables and venous blood levels of lactic acid, K+, TBARS, RAA and HSPs. The healthy volunteers in the control group underwent the same protocol.

Table 2.   Resting values of biochemical variables in controls and chronic fatigue syndrome (CFS) patients
 Glucose (mmol L−1)Cholesterol (mmol L−1)Potassium (mmol L−1)Sodium (mmol L−1)LA (mmol L−1)IL-1β (pg mL−1)TBARS (nmol mL−1)RAA (nmol mL−1)HSP27 (ng mL−1)HSP70 (ng mL−1)
  1. LA, lactic acid; IL-1β, interleukin 1-β; TBARS, thiobarbituric acid reactive substances; RAA, reduced ascorbic acid; HSP, heat shock proteins.

  2. Values are the mean ± SEM.

  3. *< 0.05; **< 0.01; ***< 0.001, significant differences from controls.

Controls4.82 ± 0.324.62 ± 1.103.99 ± 0.08140 ± 0.41.28 ± 0.181.67 ± 0.031.38 ± 0.08144 ± 1112.3 ± 0.61.64 ± 0.18
Group I CFS No history5.05 ± 0.155.95 ± 1.653.92 ± 0.12138 ± 0.51.57 ± 0.301.72 ± 0.062.00 ± 0.70135 ± 319.3 ± 1.70.58 ± 0.16***
Group II CFS Physical hyperactivity5.20 ± 0.305.31 ± 0.603.98 ± 0.06139 ± 1.01.31 ± 0.081.74 ± 0.192.24 ± 0.39*161 ± 2210.2 ± 1.30.43 ± 0.20***
Group III CFS Infection4.71 ± 0.284.58 ± 0.333.93 ± 0.08141 ± 1.51.57 ± 0.991.60 ± 0.033.57 ± 0.62***123 ± 10*18.9 ± 6.60.48 ± 0.21***
Group IV CFS Hyperactivity plus infection4.68 ± 0.135.45 ± 0.653.92 ± 0.07140 ± 1.01.45 ± 0.101.62 ± 0.043.53 ± 0.46***120 ± 11*11.5 ± 1.70.35 ± 0.16***

Exercise protocol

During the exercise trial, heart rate was computed from standard ECG leads using a software system (Oxycon beta; Jaeger, Bunnik, the Netherlands) and data were obtained for each respiratory cycle. Blood pressure was measured after every two increases in workload. Percutaneous oxygen saturation was continuously measured throughout the exercise challenge and the recovery period using an infrared analyser (model N3000; Nellcor, Houston, TX, USA). A face mask (dead space 30 mL) was designed to form an air-tight seal over the patient’s nose and mouth, with all the inspired and expired gases going into a turbine flowmeter (Triple V digital volume transducer; Jaeger) to provide measurements of minute ventilation. A side pore of the face mask was connected to fast-response differential paramagnetic O2 and infrared CO2 analysers (90% response time in 100 ms) that measured the end-tidal partial pressures of O2 and CO2, respectively. A calibration procedure for the flowmeter and gas analyser systems was carried out before each test. Exercise trials were performed in the morning, on an electrically braked cycle ergometer (Ergometrics ER 800; Jaeger) connected to a microcomputer using Oxycon beta software. The exercise bout was preceded by a 2-min 0-W pedalling period. The load was increased as a ramp (20 W min−1) until the subjects decided to terminate exercise; then, they continued to pedal for the first 5 min of the 60-min recovery period during which venous blood was collected at 0 (end of maximal power step), 5, 10, 30 and 60 min. Throughout the incremental exercise trial, breath-by-breath ventilation and O2 and CO2 consumption (inline image and inline image) were computed (Oxycon beta). The criteria used to establish maximum inline image (inline image) were the following: obtaining a plateau of inline image, reaching the predicted maximum values of inline image and heart rate, and measuring a respiratory quotient higher than 1.1 [37].

Biochemical analyses

A 6-mL sample of heparinized blood was collected at various time-points during the protocol to measure different biochemical variables. Plasma TBARS and RAA were analysed according to previously published procedures [1, 13, 17] and based on the original methods of Uchiyama and Mihara [38] for TBARS and Maickel [39] for RAA. Plasma IL-1β was measured with a high-sensitivity enzyme-linked immunosorbent assay (ELISA) (Human Quantikine Immunoassay; R&D Systems Europe, Lille, France). The limit of detection of the IL-1β assay was <1 pg mL−1. Plasma HSP27 and HSP70 levels were measured with high-sensitivity ELISAs (Human Hsp27 Total; BioSource International, Inc., Camarillo, CA, USA, supplied by Invitrogen, Eragny sur Oise, France and kit from Assay Designs, supplied by Tebu-Bio SAS, Le Perray en Yvelines, France, respectively). The limits of detection of the HSP27 and HSP70 assays were <0.3 and <0.09 ng mL−1, respectively. All measurements were made in duplicate, and the coefficient of variation was inferior to 5%.

Statistical analyses

Data are presented as means ± one standard error of the mean (SEM). To determine the magnitude and time of the postexercise peak variations in biochemical variables in each group (Tables 2 and 4), we used a repeated measures anova when the variables were normally distributed or the Friedman’s test for repeated measures when they were not. When the data were normally distributed, a two-way analysis of variance for repeated measures (RM-anova 2) (i.e. groups and time) was used to determine significant intergroup differences between the resting biochemical levels (Fig. 1) and the exercise-induced response (Fig. 4). Post hoc tests with Bonferroni correction were used to compare specific differences between groups. In all cases, significance was set at the 0.05 level. We determined the existence of least square regressions between the history of physical activity prior to CFS symptoms and the resting levels of HSP27 and HSP70 (Fig. 2), and also between the resting levels of HSP27 and HSP70 (Fig. 3).

image

Figure 1.  Resting plasma levels of a marker of lipoperoxidation [thiobarbituric acid reactive substance (TBARS)], an antioxidant [reduced ascorbic acid (RAA)] and heat shock proteins (HSP27 and HSP70) in control healthy subjects and four groups of chronic fatigue syndrome (CFS) patients: no relevant history (group I), high-level sport practice (group II), acute and severe infection (group III) or combination of physical activity and infection (group IV). *P < 0.05; ***P < 0.001, significant difference from controls.

Download figure to PowerPoint

image

Figure 2.  Correlations between the history of physical activity prior to chronic fatigue syndrome symptoms and resting levels of HSP27 and HSP70. Regression lines with 95% confidence intervals (dashed lines) are shown.

Download figure to PowerPoint

image

Figure 3.  Correlation between resting levels of HSP27 and HSP70. Regression line and 95% confidence intervals (dashed lines) are shown.

Download figure to PowerPoint

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Resting levels of variables

Resting concentrations of glucose, cholesterol, electrolytes, lactic acid and IL-1β did not differ between CFS patients and control subjects (Table 2). By contrast, the groups II, III, and IV CFS patients had significantly higher TBARS and lower RAA levels compared with controls (Table 2). The level of HSP27 did not differ between patients and controls, whereas the mean HSP70 level was lowered in all CFS groups. Figure 1 highlights the intergroup differences. In our CFS population, the resting plasma levels of HSP27 and HSP70 were negatively correlated with the history of physical activity (expressed in h per year) as shown in Fig. 2. We also noted that the resting HSP27 and HSP70 levels were positively correlated (Fig. 3). There was no correlation between the duration of CFS symptoms prior to examination and the resting levels of TBARS, RAA, HSP27 or HSP70.

Response to maximal cycling exercise

Maximal exercise performance did not significantly differ between CFS patients and controls (Table 3). Table 4 shows the maximum postexercise variations in lactic acid, oxidant–antioxidant status and HSPs measured in controls and the four CFS groups. The postexercise changes were also related to the corresponding inline image to eliminate any artefact because of the modest interindividual scattering of metabolic performance. We did not find any significant intergroup difference between the maximal lactic acid variations, which always occurred 5 min after the end of exercise. On the other hand, the peak TBARS increase occurred earlier and was significantly elevated in all CFS patients, with the maximal postexercise TBARS changes increasing significantly in CFS groups II, III and IV. In parallel, RAA consumption was delayed in groups III and IV, and the maximal postexercise RAA variations were significantly reduced in group IV patients. Table 4 clearly shows a reduction in the postexercise HSP27 and HSP70 concentrations in patients in groups II, III and IV. We noted that the postexercise HSP27 variations were significantly delayed, whereas the changes in HSP70 occurred earlier in all patients, compared with controls. Figure 4 highlights the intergroup differences showing the highest postexercise reduction in the HSP response in patients in groups III and IV. We found no correlation between the resting levels of TBARS, RAA and HSPs and their maximum postexercise variations.

Table 3.   Maximal exercise performances in controls and chronic fatigue syndrome (CFS) patients
 Ventilatory threshold (mL STPDO2 min−1 kg−1)inline image (mL STPDO2 min−1 kg−1)Pmax (W)Exercise duration (min)HR (per min)
RestMaximum
  1. inline image, maximal oxygen uptake; Pmax, maximal power; HR, heart rate.

  2. Values are the mean ± SEM.

Controls21.2 ± 1.026.7 ± 1.1170 ± 108 ± 276 ± 3156 ± 5
Group I CFS No history19.0 ± 2.022.1 ± 2.4120 ± 136 ± 295 ± 716 ± 8
Group II CFS Physical hyperactivity20.0 ± 2.029.0 ± 2.0173 ± 148 ± 378 ± 4153 ± 5
Group III CFS Infection19.5 ± 3.128.1 ± 3.0169 ± 179 ± 486 ± 6162 ± 7
Group IV CF Hyperactivity plus infection18.2 ± 2.327.1 ± 2.4158 ± 188 ± 278 ± 3148 ± 6
Table 4.   Exercise-induced variations in plasma levels of lactic acid (LA), thiobarbituric acid reactive substances (TBARS), reduced ascorbic acid (RAA) and heat shock proteins (HSP27 and HSP70) in patients with chronic fatigue syndrome (CFS)
 ControlsNo history Group I CFSPhysical hyperactivity Group II CFSInfection Group III CFSHyperactivity plus infection Group IV CFS
  1. Maximal post-exercise variation (peak), the difference between maximal and resting values (absolute and related to the corresponding inline image) and the time of peak variation are shown.

  2. Values are the mean ± SEM.

  3. *< 0.05; **< 0.01; ***< 0.001, significant difference from control subjects.

Lactic acid (mmol L−1)
 Peak6.07 ± 0.614.73 ± 0.505.95 ± 0.636.16 ± 0.585.17 ± 0.87
 ΔLAmax+4.79 ± 0.59+3.43 ± 0.43+4.64 ± 0.55+4.73 ± 0.61+3.67 ± 0.83
 inline image+0.18 ± 0.01+0.16 ± 0.02+0.16 ± 0.03+0.17 ± 0.01+0.14 ± 0.01
 Time (min)55555
TBARS (nmol mL−1)
 Peak2.19 ± 0.155.40 ± 0.90**5.57 ± 1.10*7.03 ± 1.54**8.22 ± 1.11***
 ΔTBARSmax+0.81 ± 0.2+3.40 ± 2.9+3.33 ± 1.0*+3.46 ± 1.5*+4.70 ± 1.8*
 inline image+0.03 ± 0.01+0.15 ± 0.08+0.11 ± 0.04*+0.12 ± 0.05*+0.17 ± 0.06*
 Time (min)10 ± 24 ± 1*5 ± 2*4 ± 2*5 ± 2*
RAA (nmol mL−1)
 Peak97 ± 689 ± 32103 ± 2189 ± 10101 ± 35
 ΔRAAmax−47 ± 10−46 ± 7−58 ± 17−34 ± 18−19 ± 8*
 inline image−1.76 ± 0.20+2.08 ± 0.60−2.00 ± 0.42−1.20 ± 0.51−0.70 ± 0.48*
 Time (min)0 ± 35 ± 34 ± 110 ± 5*6 ± 2*
HSP27 (ng mL−1)
 Peak25.2 ± 2.413.1 ± 3.0**14.1 ± 4.012.5 ± 6.6*9.8 ± 0.9***
 ΔHSP27max+12.7 ± 2.8+3.8 ± 4.2+3.9 ± 2.8*+0.25 ± 3.3**−1.7 ± 1.9***
 inline image+0.48 ± 0.09+0.26 ± 0.15+0.08 ± 0.10**+0.009 ± 0.002***−0.063 ± 0.008***
 Time (min)0 ± 210 ± 5*9 ± 4*15 ± 3**12 ± 4*
HSP70 (ng mL−1)
 Peak2.21 ± 0.300.90 ± 0.300.52 ± 0.250.34 ± 0.080.06 ± 0.09
 ΔHSP70max+0.57 ± 0.15+0.32 ± 0.12+0.09 ± 0.16*−0.14 ± 0.09***−0.29 ± 0.20***
 inline image+0.021 ± 0.007+0.014 ± 0.010+0.003 ± 0.006*−0.023 ± 0.009***−0.011 ± 0.008***
 Time (min)35 ± 530 ± 4**25 ± 5***28 ± 8**23 ± 5**
image

Figure 4.  Maximal exercise-induced variations of a marker of lipoperoxidation [thiobarbituric acid reactive substance (TBARS)], an antioxidant [reduced ascorbic acid (RAA)] and heat shock proteins (HSP27 and HSP70) in control healthy subjects, and in chronic fatigue syndrome (CFS) patients in the four groups (see text). *P < 0.05; **P < 0.01; ***P < 0.001, significant difference from controls. $P < 0.05, significant intergroup differences.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

The results of the present study suggest a relationship between the history of two stress factors (a high level of physical activity and severe infection) and the levels of oxidative stress and two HSPs. However, we and others [1, 17, 20, 21, 23, 27, 40] have not previously related the baseline levels of these plasma biochemical variables and their response to exercise to any potential causes of CFS. We showed that both the resting plasma levels of markers of oxidant–antioxidant status and of exercise-induced oxidative stress were elevated in CFS patients who reported a history of a high level of sport practice and/or severe infection. At rest, the mean level of HSPs cannot serve as markers of risk of CFS. Indeed, the mean level of HSP27 did not vary, and the mean HSP70 level was markedly reduced in all CFS groups. However, we demonstrated a negative correlation between the yearly duration of sport practice and resting values of HSP27 and HSP70. By contrast, the exercise-induced HSP response markedly varied amongst the different CFS groups, with the highest postexercise reduction in HSP27 and HSP70 responses being measured in patients in groups III and IV (i.e. when a severe infection alone or accompanying intense physical activity was reported).

Methodological considerations

There may be some limitations in this study with regard to the validity of comparing individuals with a history of different fitness levels. First, changes in oxidant–antioxidant status and HSP levels might result from deconditioning in CFS patients. Secondly, the physical activity level was self-reported and could have been overestimated by patients. Thirdly, the benefits of sports training in some CFS patients might have persisted. It is well known that deconditioning reduces the antioxidant capacity and increases oxidative stress in patients with chronic respiratory insufficiency [41] as well as in animal models of hindlimb suspension [42]. However, deconditioning can be ruled out in our study because (i) it induces carbohydrate and lipid disorders [43] that were not observed during routine biochemical check-up in these CFS patients, (ii) CFS patients did not have reduced maximal exercise performance or an accentuated lactic acid response and (iii) we found no correlation between the duration of CFS symptoms (i.e. the period of suppression or reduction in the sport training programme which may lead to deconditioning) and the resting levels of oxidant–antioxidant status and HSPs. It is unlikely that the history of physical activity was misjudged because we found negative correlations between the yearly duration of sport practice and the actual resting HSP27 and HSP70 levels. In addition, CFS was diagnosed later in life in well-trained subjects because patients were examined several years after the initiation of symptoms (Table 1). During this period, patients had stopped or markedly reduced their sport practice and can be considered to have become sedentary.

Innate immunity of CFS patients

We noted that the resting IL-1β plasma level of CFS patients did not differ from that in controls. This confirms our previous observation of normal resting IL-6 and tumour necrosis factor α levels in CFS patients with the absence of an accentuated interleukin response to exercise [1]. Several studies also failed to demonstrate abnormal plasma cytokine levels in CFS patients at rest compared with healthy subjects [44, 45], whereas others reported an elevation of plasma levels of IL-6 and IL-1α [31, 33–35]. It is noteworthy that no clear evidence of a link between abnormal immunity and CFS was established [10]. Our data show a clear dissociation between the oxidant–antioxidant status and innate immunity.

Oxidant–antioxidant status of CFS patients

Oxidative stress refers to an imbalance in the pro- and antioxidant status in favour of the former. This phenomenon is highly expressed in skeletal muscles because of poor antioxidant defences [46], and the occurrence of exercise-induced oxidative stress is well documented in healthy sedentary subjects [13, 24, 47–50]. Elevated oxygen uptake during exercise is a major source of production of ROS as a result of mitochondrial activity [51]. Antioxidants play a key role in counterbalancing excessive ROS production. In CFS patients, the expression of genes of importance for mitochondrial activity and oxidative balance, including superoxide dismutase, is altered [52]. This might explain the enhanced plasma TBARS levels measured in these patients in the present study, as in previous studies [1, 17–19]. However, altered gene expression in CFS does not seem to affect aerobic and anaerobic metabolism. Indeed, as we found previously [17], maximal oxygen uptake was not reduced and lactic acid production was not enhanced after maximal exercise. This agrees with most previous observations that have shown a lack of increased lactic acid production in the majority of CFS patients [53, 54]. Thus, measurements of maximal inline image and blood lactic acid changes in response to exercise cannot serve to identify those patients with CFS.

Heat shock proteins in CFS patients

In healthy subjects, acute interrelationships exist between the redox status and HSP expression. Numerous data suggest that HSPs protect the cells against the deleterious effects of ROS. Both HSP70 and HSP72 reduce oxidative stress via the activation of antioxidants [2], whereas HSP27 seems to directly scavenge free radicals [55] and/or to protect the constitutive cell proteins against the toxicity of ROS [56]. Both ROS and antioxidants regulate HSP expression, and, in turn, a reduction in HSP expression increases the generation of ROS [2]. HSP expression markedly differs between red and white muscle fibres, with a lower level reported in the latter in porcine muscle [57, 58]. These animal observations could explain the reduced baseline levels of HSPs and their attenuated response to exercise in CFS patients in the present study. Indeed, Pietrangelo et al. [59] have clearly shown an increased percentage of fast, white muscle fibres in CFS patients. In addition to these findings in different muscle types, several human studies have demonstrated an attenuated HSP response to stress in elderly individuals in relation to enhanced oxidative stress [60, 61]. It is tempting to speculate that the accumulation of two or more stressors might be the cause of a lowering of spontaneous and induced HSP production. Such a phenomenon might be responsible for the reduced baseline levels of HSPs and their response to exercise in individuals who have practised high-level sport for several years and have experienced severe infection. Because it is well documented that exercise training at a high level favours respiratory infection and reduces global immune function [62], the incidence of severe infection in CFS patients with a history of intense sport practice is not surprising.

Implication of this study

The results of the present study show that measurements of plasma HSP concentration at rest cannot be used to differentiate between CFS groups. However, because resting HSP levels are inversely correlated with the importance of physical activity, general practitioners and, in particular, sports medicine specialists should be aware of the occurrence of reduced resting HSP levels in athletes who participate in high-level training programmes and present with symptoms of CFS. The response to exercise is better for the diagnosis of CFS and allows risk factors to be identified. Indeed, our data indicate that the combination of increased exercise-induced oxidative stress and a reduced HSP response may constitute a specific marker for the diagnosis of CFS. Because the highest postexercise reduction in HSP27 and HSP70 concentrations was measured in CFS patients reporting an episode of severe infection or the combination of infection and sport practice, a maximal exercise test in these CFS patients seems highly relevant.

In conclusion, the results of the present study suggest that the combination of two or more stressors might be the cause in later life of a persisting depletion of HSP production resulting in an exacerbation of oxidative stress.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

The authors gratefully acknowledge Dr Chantal Guillot for her helpful participation in this study.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References