INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE CHRONIC FATIGUE SYNDROME (CFS) is a clinically defined condition characterized by severe and unexplained fatigue. The Centers for Disease Control definition of CFS requires that a 50% reduction in activity level be accompanied by other documented somatic complaints (12). However, the pathophysiological basis of the fatigue and exercise intolerance is not well understood.

Several approaches have been used to investigate potential physiological alterations in CFS. Much of this work has attempted to identify peripheral defects in muscle metabolism. For example, there have been reports of reduced oxidative metabolism (26, 32), oxidative damage to muscle (13), mitochondrial abnormalities (2), increased lactate production (22), and reduced oxygen delivery (25). Others have reported that muscle metabolism is normal (3, 19). This has lead some to suggest that the reported physiological abnormalities are a consequence of extreme deconditioning and not due to any specific organic cause of CFS (31).

The relationship between altered central hemodynamics and fatigue has been investigated less extensively. Hypovolemia, which is documented in some individuals with CFS (30), is a potential mechanism that may contribute to the fatigue and exercise intolerance noted by these patients. Chronic or acute hypovolemia could lower peak oxygen consumption (O₂) and cause exercise intolerance by reducing left ventricular stroke volume and cardiac output. Support for this concept comes from prior studies in healthy young and older adults where strong relationships among blood volume, maximal O₂, and physical activity were reported (5, 15, 16, 18). Thus a potential vicious cycle exists with fatigue and the consequent decrease in physical activity leading to reductions in blood volume, stroke volume, and peak O₂. The initiating factors in this vicious cycle are unknown.

We therefore investigated the relationships among supine blood volume, stroke volume, and peak oxygen uptake in patients with CFS. We hypothesized that patients with CFS would have lower blood volume compared with controls. Furthermore, we hypothesized that the fundamental relation between blood volume and maximal O₂ reported in healthy subjects would also hold for patients with CFS, supporting the concept that volume status is a strong physiological correlate of exercise tolerance. Because habitual level of physical activity can influence blood volume and peak O₂ (18), we examined the influence of physical activity on these variables. We hypothesized that habitual physical activity would be strongly correlated with blood volume and peak O₂. Finally, because fatigue is the major complaint noted by this patient group, we attempted to identify physiological variables that might predict the severity of fatigue. We hypothesized that fatigue would be inversely related to blood volume and peak O₂.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Seventeen patients and 17 healthy control subjects gave verbal and written consent to participate in this institutionally approved study. All patients fulfilled the Fukuda et al. (12) criteria for a diagnosis of CFS. Patients and controls provided a full medical history and had a physical examination. Patients and control subjects were required to be free from any acute illness or chronic disease. All participants completed the College Alumnus Health Questionnaire (23) to assess weekly physical activity. Physical activities recorded on the questionnaire were classified according to energy expenditure (1). The fatigue-severity scale questionnaire was used to quantify level of fatigue (20). General physical well being (physical component score) was also assessed in all subjects with the Medical Outcome Study short-form general health survey (29). The physical component score was used as an additional surrogate measure of fatigue. Body composition was assessed using skinfold calipers (17).

Peak O₂. Peak O₂ was determined via open-circuit spirometry with a ParvoMedics Truemax 2400 Metabolic Measurement System (Consentius Technologies; Sandy, UT) during cycling exercise to volitional fatigue on a Monark upright cycle. The protocol consisted of 2-min stages where workload was increased by 25-30 watts per stage. O₂, ventilation, CO₂ production, and the respiratory exchange ratio (R values) were recorded throughout the test and grouped in periods of 30 s. Blood pressure and heart rate (3-lead electrocardoigram) were also recorded during the protocol. All subjects were encouraged to give maximal efforts; in addition, an R value of >1.0 was required for a test to be considered a peak effort.

Plasma and blood volume determination. Plasma volume was determined using a single bolus injection (3.0-3.5 ml) of Evans blue dye (concentration 5 mg/ml; New World Trading; Debary, FL) following a 30-min rest in the supine position and a 12-h fast. Water (5 ml/kg) was given to all subjects 1 h before the procedure. A previously placed antecubital vein catheter was used for the injection and subsequent blood sampling. Care was taken to adequately flush the catheter with saline following the Evans blue injection. Using the same catheter for injection and blood sampling has been previously performed without significant loss of accuracy (9). The absorbance of the plasma samples was read at 620 nm 10, 20, and 30 min after the injection with the use of a direct spectrophotometric technique (Beckman Spectrophotometer, Beckman Instruments; Fullerton, CA) (10). Reported values are based on the peak absorbance reading, which occurred at 10 min in most studies. Plasma volume (PV) was calculated using the following formula: PV = 20 · (ml of dye injected/4) · (absorbance of standard/dye concentration)/(absorbance of plasma sample). Others (6, 14) have shown that calculation of plasma volume by this method yields nearly identical results to plasma volume determination from an extrapolation to time 0 of the dye disappearance curve. This method is also highly reproducible. For example, plasma volume was measured in four healthy volunteers on two separate occasions and found to be 43.7 ml/kg on the first occasion and 44.1 ml/kg on the second session. The coefficient of variation was <4%. This high reproducibility has also been reported by others (6, 14). Hematocrit (Hct) was determined in duplicate or triplicate using a microcentrifuge and corrected (Hct_corr) for peripheral sampling (0.91) and trapped plasma (0.96). Blood volume (BV) was calculated using the formula BV = PV/(1  Hct_corr). There are some limitations to using this approach. For example, loss of dye from the vascular space after injection would result in an overestimation of plasma and blood volumes. However, although the peak absorbance usually occurred at 10 min postinjection, the 20- and 30-min samples were only marginally lower than the peak sample, indicating minimal loss of dye from the vascular space in the short time period of the procedure (<30 min). In addition, identical procedures were utilized for all studies, thereby making any possible overestimation of plasma volume similar between patients and controls.

Baseline cardiac output and stroke volume. The indirect Fick method of CO₂ rebreathing (4) was used to derive supine cardiac output (Q_c) from which stroke volume (using heart rate) was calculated. During steady-state breathing, CO₂ production (VCO₂) and end-tidal CO₂ were measured. End-tidal CO₂ provided an estimate of arterial CO₂ (Ca_CO2). To estimate venous CO₂ (Cv_CO2), subjects rebreathed a high CO₂-O₂ gas mixture from a rebreathing bag until the level of CO₂ in the bag and lung reached equilibrium. This equilibrium value for CO₂ provided an estimate of Cv_CO2. The following equation was then used to calculate cardiac output: Q_c = VCO₂ / Cv_CO2  Ca_CO2.

Statistics. Data are reported as means ± SE. Comparisons between patients and controls were made using two-tailed unpaired t-tests. Univariate correlation coefficients (Pearson) between variables of interest were examined. A multivariate, stepwise regression model was constructed with peak O₂ as the dependent variable and blood volume, stroke volume, physical activity, and fatigue as the independent variables, with forward entry and removal (Systat, SPSS). Another regression model was constructed with fatigue as the dependent variable and peak O₂, blood volume, stroke volume, and physical activity as the independent variables. The level of significance was set at P  0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject demographic data can be found in Table 1. There were no significant differences in any of the measured variables. Supine hemodynamic data can be found in Table 2. Patients had significantly faster resting heart rates (P  0.05) and higher diastolic pressures (P  0.05). There was a trend for blood volume (Fig. 1; P = 0.084) and plasma volume (39.5 ± 1.5 vs. 42.9 ± 1.7 ml/kg; P = 0.14) to be lower in patients compared with controls. Similar results were noted when blood volume was expressed relative to fat-free mass (data not shown). There was no significant difference in hematocrit between groups (36.9 ± 0.98 vs. 39.1 ± 0.67; P = 0.23). Peak O₂ was significantly lower in patients compared with controls (Fig. 2; P < 0.001). Peak heart rate during the exercise test was not different (169 ± 5 vs. 176 ± 3 beats/min; P = 0.38) nor was the percentage of age-predicted maximal heart rate (94 ± 3 vs. 96 ± 1%; P = 0.48) between groups. The R values at peak exercise were 1.16 ± 0.03 in patients and 1.20 ± 0.02 in controls (P = 0.19). Only one patient (and no controls) failed to achieve an R value of >1.0 so, therefore, these data were excluded from the analysis. Peak work rate was lower in the patients (100 ± 6 vs. 173 ± 16 watts, P  0.01). Peak ventilation was 47 ± 6 l/min in patients and 74 ± 6 l/min in controls during the exercise test (P = 0.005).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Individual and means ± SE data for blood volume expressed relative to body weight in those with chronic fatigue syndrome (CFS) (n = 17) and controls (n = 17). There was a trend for blood volume to be lower in the patients (P = 0.089).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Individual and means ± SE data for peak oxygen consumption (O2) relative to body weight in those with CFS (n = 16) and controls (n = 17). Patients with CFS had a significantly lower peak O2 compared with the controls (P < 0.001). Note that one patient was deleted from this comparison for failing to achieve an R value of >1.0 during the test.

Questionnaire data. The physical activity questionnaire indicated that patients engaged in significantly less physical activity on a weekly basis compared with controls (1,018 ± 225 vs. 5,468 ± 1,301 kcal/wk, P  0.01). The fatigue severity questionnaire indicated that patients experienced significantly more fatigue than controls (scale 1-7 units; 5.1 ± 0.18 vs. 1.3 ± 0.23 units, P < 0.0001). The PCS indicated that patients were more limited than the controls (1.58 ± 0.30 vs. 0.57 ± 0.11, P < 0.001).

Correlations and regressions. There was a strong relationship between blood volume and peak O₂ within the CFS and control groups (see Table 3 and Fig. 3). This relationship also existed when the data are expressed relative to body weight. Other key correlations among hemodynamic data, physical activity, and fatigue can be found in Table 3. With the use of multivariate stepwise regression, the relationship between blood volume and peak O₂ was not strengthened with the addition of other independent variables. Therefore, only the univariate correlations are presented. The multivariate stepwise regression using fatigue as the dependent variable indicated that there were no significant predictors.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Relationship between blood volume and peak O2 in individuals with CFS and control.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The major findings from the present investigation are as follows: 1) plasma and blood volumes are not significantly different between CFS patients and controls, although a trend for lower volumes was apparent in the patients; 2) peak O₂ and peak ventilation were significantly lower in the CFS group compared with age-matched controls; 3) the fundamental relationship between blood volume and peak O₂ found previously in healthy subjects (5, 16) was confirmed and extended to those with CFS; and 4) self-reported fatigue and the PCS of the short-form general health survey did not correlate with any of the measured physiological parameters but did correlate inversely with physical activity in the CFS group.

Patients diagnosed with CFS report a variety of symptoms including severe, debilitating and unexplained fatigue that results in functional impairment. There are previous reports of both preserved (21) and impaired (27) peak O₂, an objective index of functional capacity, in this patient group. Despite these conflicting reports, we anticipated that because of the chronic fatigue, patients with CFS would have a lower peak O₂ compared with healthy controls. Our data demonstrate that patients had a 35% lower peak aerobic power compared with a group of age-matched healthy controls. One potential reason for the lower peak O₂ could be inadequate effort by the patients. However, markers of maximal effort, such as the R values at peak exercise and the percentage of age-predicted maximal heart rate achieved provide objective evidence that the patients put forth a maximal effort. The current data agree with a recent report (7) where female CFS patients and sedentary age-matched controls underwent exercise testing on a cycle ergometer. Similar to our findings, peak O₂ and work rate were significantly impaired in those patients with CFS. These data, demonstrating lower peak O₂, confirm the limited functional capacity of this patient group. Whereas peak O₂ may provide an objective measure of the functional limitations in this patient group, the underlying mechanism for this impairment is inadequately understood.

To the best of our knowledge, no studies to date have examined the relationship between blood volume and peak O₂ in patients with CFS. We hypothesized that the impaired peak aerobic power would be explained, in part, by hypovolemia. In support of this hypothesis, there was a trend for blood volume to be lower in patients. Importantly, within each subject group, there was a strong relationship between blood volume and peak O₂. This relationship may be mediated through the effects of stroke volume on central hemodynamics, because stroke volume was correlated with blood volume and peak O₂. This relationship is consistent with previous observations in young and older healthy adults (5, 15, 16).

A frequent criticism of CFS studies is the lack of a sedentary control group. Despite our desire to recruit sedentary controls, there were clear differences in levels of physical activity between the cohorts. The current data collected using a standard questionnaire (23) are in agreement with data collected using an accelerometer (28). Nevertheless, even with different levels of physical activity, the anticipated relationships between blood volume and peak O₂ were present in this study.

A potential vicious cycle exists in patients with CFS whereby the illness results in cardiovascular deconditioning that further exacerbates the symptoms of the underlying illness. For example, symptoms associated with CFS may result in reduced physical activity that leads to reduced blood volume (5) and consequently reduced peak O₂. Thus our data do not necessarily support the argument that deconditioning is a primary causative factor for CFS (31), rather they support the anticipated relationship between blood volume and peak O₂, irrespective of the cause of CFS.

The fatigue severity scale confirmed the presence of the major symptom reported by this patient group and clearly differentiated the patients from the controls. However, the patients' fatigue severity scores did not correlate with peak O₂ or blood volume. These scores, however, were tightly clustered within each group, minimizing the likelihood of establishing a relationship between fatigue and these measures. The PCS measure of the short-form general health survey demonstrated greater scatter within groups but also did not correlate with hemodynamic measures.

These results have two possible implications. First, it is possible that scales chosen were not appropriate instruments to measure the fatigue and physical health of the patients; there is no widely accepted instrument for measuring the subjective symptoms associated with CFS. Second, it is possible that there is in fact no relationship between these subjective measures of fatigue or well being and the patients' peak O₂ and blood volume. Although speculative, there may be discordance between symptom perception and physiology in this population. We (11) have previously suggested that patients with idiopathic orthostatic intolerance and chronic fatigue reported more symptoms than controls in response to a graded lower body negative pressure stress despite equivalent physiological responses. The present data lend additional support to the discordance between symptoms and physiological responses. It is thus possible that therapeutic interventions may increase peak O₂, yet not improve fatigue as measured using these scales. These considerations should be borne in mind when attempting to quantify fatigue level or when designing therapeutic end points for patients with CFS.

Although these data support the established relationship between blood volume and peak O₂ in patients with CFS, the study design does not permit us to infer a causal relationship between these variables. The data, however, do provide insight into a potential mechanism whereby reduced blood volume may result in reduced exercise tolerance. Furthermore, our data provide a testable strategy for interventions that may alter the course of this disabling illness. Also, whereas we have demonstrated a strong relationship between volume status and peak O₂, other potential central hemodynamic effects may also be responsible for the lower peak O₂. For example, we cannot exclude subtle cardiac dysfunction as has been reported in selected CFS patients (8, 24).

In summary, patients with CFS had a significantly lower peak O₂, lower peak ventilation, and a trend toward a lower blood volume. The physiological relationship between blood volume and peak O₂ is maintained in this population.