Evidence for aerobic insufficiency in women with systemic lupus erythematosus




To determine if fatigue is associated with diminished aerobic capacity in women with systemic lupus erythematosus (SLE).


Eighteen women (age 35 ± 9 years) with mild SLE (Systemic Lupus Activity Measure = 3.1 ± 2.1) and 16 healthy but sedentary controls (age 38 ± 8 years) completed peak treadmill exercise tests to determine aerobic capacity and Fatigue Severity Scales to quantify the severity of fatigue. Measures of oxygen consumption (VO2) were recorded during the treadmill tests.


Peak VO2 was lower in patients with SLE (19.2 ± 4.4 ml/kg/minute) as compared with controls (27.4 ± 4.7 ml/kg/minute) and expected values (30.7 ± 3.1 ml/kg/minute; P < 0.0006 versus controls and P < 0.0001 versus expected). Functional aerobic impairment was observed in 14 of the 18 patients with SLE. In patients with SLE, ventilatory threshold, a marker for the onset of lactic acidemia, was observed at the lowest energy requirement for instrumental activities of daily living. Peak VO2 in the patients with SLE was similar to the highest energy requirements for instrumental activities of daily living, leaving little or no reserve for more intense occupational and recreational activities. Peak VO2 was significantly higher (P < 0.0001) than the activity of daily living requirements in controls, providing a substantial energy reserve. Fatigue severity score (FSS) was 5.0 ± 1.4 in patients with SLE, with 14 of the 18 patients having scores above 4.0, a score indicating that fatigue severity limited physical activity. Of the 14, 12 had functional aerobic impairment. An FSS of greater than 4.0 was not observed in controls (mean = 2.5 ± 0.7).


In women with SLE, aerobic capacity was diminished to levels that were insufficient for engaging in activities of daily living and below those expected to result from physiologic deconditioning. This functional aerobic impairment was strongly correlated with the perception of severe, activity-limiting fatigue.


Nearly all (80%) people who have systemic lupus erythematosus (SLE) identify excessive fatigue as a major symptom that limits their quality of life (1–3) and physical functioning (4). Of these patients, 30–50% report fatigue as their most debilitating symptom. Severe fatigue is often caused by attenuated ability of the oxidative metabolic pathway to supply energy for physical activity (i.e., low aerobic capacity). Diminished aerobic capacity has been reported, in conjunction with fatigue, in patients with SLE (5–7).

Aerobic capacity is usually measured while exercising on a cycle ergometer or motor-driven treadmill. Use of the cycle ergometer has certain safety advantages and the workload can be regulated at smaller increments than those usually available with treadmill testing. Intrasubject and intersubject measurement error are greater on the cycle ergometer than on the treadmill because 1) both attainment of a maximal workload and maintenance of a submaximal workload are dependent on subjects' voluntary effort, 2) the muscle mass required for pedaling is often inadequate for inducing maximal oxidative metabolism, and 3) the resultant peak aerobic capacity, measured on the cycle ergometer, is much lower than that measured on the motorized treadmill in individuals who are not conditioned to cycling (8–10). These biases limit the usefulness of the results of cycle ergometer testing for determining patients' ability to engage in physical activity. The motor-driven treadmill is the most frequently used method of exercise testing for clinical purposes. The treadmill mechanically imposes the workload and does not require volitional effort to regulate the workload; the subject is required to continue walking at a given speed and elevation until the workload is changed or the treadmill stopped by the individuals conducting the test. Measurements of aerobic capacity obtained from treadmill testing are highest among all forms of standardized exercise testing (8–10). Results of treadmill tests have been used to establish a clinical threshold for functional aerobic impairment (FAI; defined as a peak oxygen consumption of 27% or more below expected values) (11) and normative reference values for aerobic capacity that are specific to sex and decade of life (12). Previous reports of diminished aerobic capacity in patients with SLE have been based on measurements obtained during cycle ergometer testing (6, 7). In addition to providing a more accurate assessment of maximum aerobic capacity, studies using treadmill testing would provide information suitable for determining the influence of diminished aerobic capacity on fatigue and the inability to engage in physical activity in these patients (13, 14). The purpose of this study was to determine if fatigue is associated with diminished aerobic capacity in women with mild SLE.



Subjects in this study were 18 physically inactive women with mild SLE and 16 physically inactive, but otherwise healthy, sex-matched controls with mean age and body mass that were not significantly different from the subjects with SLE (Table 1). Classification as physically inactive was similar for both the women with mild SLE and controls: operationally defined as not having participated in activities that evoke perspiration, for 10 minutes or longer, once or more per week.

Table 1. Group demographics*
  • *

    Data are means ± standard deviations. There were no significant differences between the patients with systemic lupus erythematosus (SLE) and controls for any of the descriptive variables listed. BMI = body mass index; bpm = beats per minute; VO2 = oxygen consumption.

Age, years35 ± 938 ± 8
Weight, kg72.2 ± 6.867.7 ± 14.9
BMI, kg/m226.6 ± 5.224.4 ± 5.7
Height, cm163.8 ± 9.0165.8 ± 6.8
Rest heart rate, bpm75 ± 974 ± 7
Rest VO2, ml/kg/minute2.2 ± 0.52.4 ± 0.9
 African American94

The patients with mild SLE had Systemic Lupus Activity Measure (SLAM) (15) scores averaging 3.1 (SD 2.1, range 0–9), indicating mild disease activity. All of the patients with mild SLE met the American College of Rheumatology (formerly American Rheumatism Association) 1982 revised criteria for the classification of SLE (16) at the time of diagnosis. Of the 14 patients who were taking prednisone, 11 were on dosages of 5–10 mg/day, 1 was on a dosage of 15 mg/day, 1 was on a dosage of 20 mg/day, and 1 was on a dosage of 25 mg 2 times per day. Four of the patients with SLE were not taking steroids.

None of the subjects in either group had concomitant disease manifestations or symptoms known to limit exercise performance at the time of the study. None of the subjects were taking medications that would limit or enhance exercise performance. Prior to participation, the procedures were explained to each subject and written consent to participate was obtained in accordance with the procedures approved by the University of Maryland, Baltimore, institutional review board.


Following informed consent, subjects completed the Fatigue Severity Scale (4). Subjects then rested for 10 minutes so that baseline data could be obtained. After the rest period and recording of the baseline data, exercise data were acquired using a modified Bruce treadmill protocol (17). The tests continued until volitional exhaustion, as identified by the subject's indication that she must stop exercising, despite strong encouragement to continue by the investigational team. Heart rate and metabolic data were recorded continuously.

Fatigue scale.

The Fatigue Severity Scale was used to assess perceived fatigue (4). This scale is a scientifically validated tool that assesses an individual's perception of the degree to which fatigue limits her or his physical and social functioning. The tool consists of 9 statements aimed at identifying how the individual perceives fatigue to limit her or his ability to function. The individual's responses to each of the 9 statements are graded on a 7-point Likert scale with a score of 7 indicating strong agreement, a score of 1 indicating strong disagreement, and a score of 4 indicating neutrality. The fatigue severity score (FSS) is determined by averaging the response scores over the 9 items. Therefore, an FSS of greater than 4 indicates that fatigue is perceived to limit physical and social functioning, an FSS of less than 4 indicates that fatigue is not perceived to limit physical and social functioning, and an FSS of 4 indicates uncertainty.

Maximum exercise testing and metabolic measurements.

Maximum exercise tests were carried out using a standard motorized treadmill. Peak oxygen consumption, a measure of aerobic capacity, was obtained using a metabolic cart housing zirconium cell oxygen and infrared cell carbon dioxide analyzers. Expired pulmonary volume was derived from flow-volume loops generated by a pneumotachometer. Gas analyzers were infused with a known mixture of 12% oxygen and 5% carbon dioxide for calibration of measurements made at the lower limit of oxygen consumption and upper limit of carbon dioxide expiration. A mixture of 21% oxygen and 0% carbon dioxide was used to calibrate measurements made at the upper limit of oxygen consumption and lower limit of carbon dioxide expiration. Both gas mixtures were contained within a nitrogen balance. The pneumotachometer was calibrated by injecting room air from a 3-liter gas syringe at various flow rates. Population referenced comparisons of peak oxygen consumption were made from published tables (12). Aerobic impairment index (AII) (11), a quantitative measure of FAI, was calculated using the standard formula:

equation image

Expected peak oxygen consumption was determined in milliliters of oxygen consumed per kilogram of body weight per minute (ml/kg/minute) using the formula 42.3 − (0.356 × age in years) for sedentary females (11). Interpretation of AII scores was as follows: <26% = no determinable FAI; 27–40% = mild impairment; 41–54% = moderate impairment; 55–68% = marked impairment; and >68% = extreme impairment. Ventilatory threshold, a marker for the onset of exercise-induced fatigue, was determined by the V-slope method (18). Heart rate was determined by continuous electrocardiogram (EKG). Expected peak heart rate was determined as 220 beats per minute minus the subject's age in years. Respiratory exchange ratio was determined as the ratio of peak carbon dioxide expiration to peak oxygen consumption. The respiratory exchange ratio expresses the relationship of anaerobic to aerobic metabolism during exercise and is frequently used as a marker to identify maximal effort. Oxygen pulse, a measure of cardiorespiratory function, was determined as the quotient of oxygen consumption divided by the heart rate and reported in ml/kg/beat. Treadmill test duration and peak test stage were functions of the time exercise was sustained on a standard test protocol, with sequential increments in workload. Workload was quantified as a unit of power output (watts), representing the work-over-time requirement for the treadmill speed and elevation.

Variables and statistics.

Two-way analyses of variance (ANOVAs) were used to identify significant intra- and intergroup differences in the dependent variables: peak oxygen consumption (VO2), AII, and peak heart rate. The first main effect analyses used a repeated-measures procedure to identify significant differences associated with independent variable classes, which were the expected and measured conditions. The second main effect analyses used an unequal-group procedure to determine significant differences associated with independent variable classes, which were the SLE and control groups.

Peak VO2 and ventilatory threshold were compared with energy requirements for activities of daily living, using repeated-measures ANOVAs. Significant intergroup differences among the dependent variables (peak respiratory exchange ratio, peak oxygen pulse, test duration, test stage, and workload) were determined by independent t-tests. In the protocol used in this study, treadmill test stages consisted of standard increments in workload, each 3 minutes. Therefore, the test stage measurements provided interval data appropriate for a parametric statistical analysis.

Spearman correlation coefficients were calculated to determine if peak VO2, AII, ventilatory threshold, test duration, test time to the ventilatory threshold, peak test stage, test stage at the ventilatory threshold, peak workload, and workload at the ventilatory threshold were related to FSS. Intergroup differences in FSS were assessed using a Wilcoxon ranked sum analysis.

For all analyses, statistical significance was assigned when type I error probability was equal to or less than 5% (P ≤ 0.05). Data are reported as means ± 1 standard deviation throughout the text.


Functional aerobic impairment.

All subjects reached the designated test end point of volitional exhaustion and were without chest pain, EKG abnormalities, dyspnea, muscular or skeletal pain, or other confounding symptoms. Peak VO2 (Table 2) was significantly (P ≤ 0.0006) lower in the patients with SLE than in the controls. When compared with reference ranges obtained from disease-free populations, peak VO2 for the control group (Table 3) was indicative of individuals with physiologic deconditioning, whereas those with mild SLE had scores that were lower than the first percentile for their decade of life and sex (Table 4). In patients with mild SLE, AII was higher than the 26% threshold for determining FAI, a finding not observed in the control group. Fourteen of the 18 patients with mild SLE (78%) were determined to have FAI: 8 with mild severity and 6 with moderate severity (Table 4). In contrast, only 2 subjects in the control group (13%) had AII scores high enough to indicate FAI: in both cases, AII in these subjects indicated mild FAI.

Table 2. Peak heart rate, oxygen uptake, and AII threshold*
  • *

    Data are means ± standard deviations. Peak oxygen uptake was also significantly lower than the expected value in the subjects with systemic lupus erythematosus (SLE) (P ≤ 0.0001). Aerobic impairment index (AII) threshold is the oxygen consumption level equal to 27% below the expected value. In the control group, peak oxygen uptake was significantly (P ≤ 0.0001) higher than functional aerobic impairment threshold. HR = heart rate; bpm = beats per minute; NS = not significant; VO2 = oxygen consumption.

Peak HRExpected, bpm185 ± 09183 ± 06NS
Peak HRMeasured, bpm172 ± 32178 ± 10NS
Peak VO2Expected, ml/kg/minute29.9 ± 3.129.1 ± 2.1NS
Peak VO2Measured, ml/kg/minute19.2 ± 4.427.4 ± 4.70.0006
AII threshold, ml/kg/minute21.8 ± 2.021.3 ± 1.6NS
Table 3. FAI severity, peak VO2 percentile ranking, and aerobic (VO2) status for control group*
SubjectAge, yearsPeak VO2, ml/kg/minuteAII, %FAI severityPeak VO2 rankingVO2 status
  • *

    A negative AII indicates measured peak VO2 was above expected value. A negative FAI index value indicates that peak VO2 was above the expected value. Peak VO2 percentile ranking and aerobic status are from reference 12. FAI = functional aerobic impairment; VO2 = oxygen consumption; AII = aerobic impairment index.

14427.1−2No FAI>20Poor
24028.4−1No FAI>30Poor
34221.521No FAI>1Very poor
44029.6−5No FAI>40Fair
54320.126No FAI<1Very poor
63823.817No FAI>1Very poor
73725.612No FAI>10Poor
84030.8−10No FAI>45Good
93827.93No FAI>15Poor
103033.0−4No FAI>45Fair
114534.7−32No FAI>85Superior
123825.312No FAI>5Poor
132223.831Mild>1Very poor
143337.0−21No FAI>70Excellent
153526.113No FAI>5Poor
162723.528Mild>1Very poor
Mean37 ± 627.4 ± 4.75 ± 17No FAI>10Very poor
Table 4. FAI severity, peak VO2 percentile ranking, and aerobic (VO2) status for subjects with SLE*
SubjectAge, yearsPeak VO2, ml/kg/minuteAII, %FAI severityPeak VO2 percentileVO2 status
  • *

    Negative AII indicates measured peak VO2 was above expected value. Peak VO2 percentile ranking and aerobic status are from reference 12.

  • FAI = functional aerobic impairment; VO2 = oxygen consumption; SLE = systemic lupus erythematosus; AII = aerobic impairment index.

13517.043Moderate<1Very poor
23720.130Mild<1Very poor
33218.540Mild<1Very poor
42316.851Moderate<1Very poor
53815.646Moderate<1Very poor
62320.340Mild<1Very poor
73721.726No FAI<1Very poor
82627.816No FAI<10Poor
94715.539Mild<1Very poor
105315.733Mild>1Very poor
112528.415No FAI>10Poor
123614.053Moderate<1Very poor
132116.951Moderate<1Very poor
143514.651Moderate<1Very poor
153918.933Mild<1Very poor
164326.13No FAI>15Poor
174316.838Mild<1Very poor
183420.731Mild<1Very poor
Mean35 ± 919.2 ± 4.336 ± 14Mild<1Very poor

Ability to engage in physical activity.

The oxygen uptake (13), or energy (14), requirements for the instrumental activities of daily living range from 10.5 ml/kg/minute to 17.5 ml/kg/minute. Peak VO2 was significantly (P < 0.0005) higher than the upper limit of the range associated with activities of daily living in the control group, but not in those with mild SLE (Figure 1). Ventilatory threshold was significantly (P < 0.0001) higher than the lower limit of the range of oxygen consumption levels associated with activities of daily living in the control group, but was similar to the lower limit for these activities in the patients with mild SLE (Figure 1). Peak heart rate exceeded 90% of the expected values in both groups (Table 2). Peak respiratory exchange ratio was 1.18 ± 0.10 in the subjects with mild SLE and 1.24 ± 0.09 in the control group. The attainment of a respiratory exchange ratio higher than 1.10 and a peak heart rate greater than 90% of the expected value indicated that both groups approached maximal aerobic capacity at volitional exhaustion. Peak oxygen-pulse was significantly (P < 0.0004) lower in the women with SLE (0.11 ± 0.02 ml/kg/beat) than in the controls (0.15 ± 0.03 ml/kg/beat). As an ANOVA main effect, treadmill test duration was shorter (P < 0.0001), fewer test stages were completed (P < 0.0001), and workload was lower (P < 0.0008) at both peak exercise and ventilatory threshold in subjects with SLE, compared with controls (Table 5).

Figure 1.

Oxygen consumption (VO2) at the ventilatory threshold and peak exercise. Error bars show means ± standard deviations. *Significantly higher than requirements for lowest activities of daily living (ADL) intensities (P < 0.0001). **Significantly higher than requirements for highest ADL intensities (P < 0.0005).

Table 5. Physical performance at peak exercise and anaerobic threshold*
  • *

    Values are means ± standard deviations. P values are for intergroup main effects. Significant interactions were not observed. SLE = systemic lupus erythematosus.

Peak exercise   
 Test stage4.61 ± 1.145.88 ± 0.620.0001
 Time, minutes12.56 ± 3.4716.24 ± 1.600.0001
 Workload, watts111.12 ± 46.57166.30 ± 49.650.0008
Ventilatory threshold   
 Test stage2.25 ± 0.653.38 ± 0.720.0001
 Time, minutes5.87 ± 2.958.40 ± 1.880.0001
 Workload, watts33.79 ± 15.5560.54 ± 32.450.0008

Self-reported fatigue.

FSS correlated indirectly with peak VO2 (r = −0.59, P < 0.0008) and ventilatory threshold (r = −0.55, P < 0.0002), and directly with AII (r = 0.66, P < 0.0001) when data were collapsed across groups (FSS available in 18 subjects with SLE and 13 controls). These variables were not related when correlation coefficients were computed for the mild SLE and control groups separately. Patients with mild SLE (FSS = 5.0 ± 1.4; n = 18) had significantly higher (P < 0.01) FSS than controls (FSS = 2.5 ± 0.7; n = 13). FSS was higher than 4, a level at which fatigue is considered to be activity limiting, in 14 of the 18 patients with mild SLE: 12 of these 14 patients were determined to have FAI (Table 6). No subject in the control group (n = 13) had an FSS higher than 4 (highest control score = 3.7). FSS was also indirectly related to test duration (r = −0.51, P < 0.0032), time to ventilatory threshold (r = −0.37, P < 0.0457), peak test stage (r = −0.47, P < 0.0227), test stage at which the ventilatory threshold occurred (r = −0.49, P < 0.0056), and peak workload (r = −0.30, P < 0.0442).

Table 6. Fatigue severity scale and aerobic impairment index in women with SLE*
ScoreNo. women with SLEPercentage of patients with SLE
  • *

    SLE = systemic lupus erythematosus; AII = aerobic impairment index; FSS = Fatigue Severity Scale score.

AII ≥ 27 and FSS > 41267
AII ≥ 27 and FSS ≤ 40211
AII < 27 and FSS > 40211
AII < 27 and FSS ≤ 40211


Nearly all patients with SLE experience significant fatigue (1–3). The Fatigue Severity Scale has been used to subjectively evaluate fatigue in SLE and other chronic conditions (1, 4, 19, 20). Previous studies reported FSS in patients with SLE ranging from 4.6 to 5.3 (1, 19). These scores were similar to the scores of patients with mild SLE in the current study. Scores of lower than 4.0 are interpreted to indicate that fatigue is not severe enough to limit participation in routine physical activity. Scores of greater than 4.0 are interpreted to indicate that fatigue is perceived to adversely affect the ability to engage in physical and social activity, with the severity of the limitation increasing with the FSS. Seventy-eight percent of the patients in the current study had FSS of greater than 4.0, despite having only mild SLE. Of the patients with FSS higher than 4.0, 86% had FAI. FSS were directly correlated with aerobic impairment index and indirectly correlated with peak oxygen consumption, ventilatory threshold, and measures of treadmill work capacity, identifying a relationship between subjective fatigue and physiologic deficits in aerobic capacity. The 2 patients who did not report FSS higher than 4.0 (n = 4) but who had FAI reported that they experienced fatigue when engaging in activities requiring energy expenditures of approximately 3.5 metabolic equivalents (METS) or greater (Figure 2). These patients did not indicate the low FSS meant that fatigue did not limit their ability to participate in daily physical activity. Rather, they had responded to the condition by accepting the limitation and adjusting their lifestyle to regularly participate only in activities they knew would not cause them to become fatigued. Including the 14 subjects with FSS greater than 4.0 (12 of these had FAI) and the 2 subjects with FSS greater than 4.0 and no FAI, 16 of the 18 (89%) subjects with mild SLE were identified as having fatigue that limited their activity. In the 2 subjects who reported FSS of less than 4.0 and who did not have FAI, fatigue did not appear to limit physical activity to a degree that was different from the sedentary, but otherwise healthy, controls. The results of this study demonstrate that, as perceived by patients with mild SLE, fatigue is clearly related to, and possibly caused by, diminished aerobic capacity.

Figure 2.

Ability of women with systemic lupus erythematosus (SLE) and controls to engage in selected physical activities (13). Bars on right indicate capacity to engage in activities without fatigue. As bars become darker, ability to engage in activity without fatigue increases. Areas without shading represent negligible ability. (1.0 metabolic equivalent [MET] = 3.5 ml/kg/minute.)

Physiologic fatigue occurs when the oxygen demand exceeds the ventilatory threshold (14, 18). Sustaining activities above the ventilatory threshold leads to exhaustion. The time from the onset of fatigue to the occurrence of exhaustion is inversely proportional to the energy required to perform an activity. Exhaustion can occur either, by definition, at peak oxygen consumption or with prolonged activity above the ventilatory threshold but below the intensity that induces peak oxygen consumption. Most activities of daily living fall within the range of energy requirements of 10.5–17.5 ml/kg/minute, and many nonsedentary occupational and recreational activities are above 17.5 ml/kg/minute (13, 14). Aerobic insufficiency is observed when ventilatory threshold and peak oxygen consumption are inadequate for avoiding fatigue while engaging in routine daily tasks.

Specific energy expenditure requirements for performing daily living, occupational, and recreational activities have been well defined (13, 14). Patients with SLE frequently complain of fatigue during routine daily activities. Results of the current study demonstrate a physiologic basis for this activity-related fatigue, which was made evident by aerobic capacity that was too low to effectively engage in activities that require more than minimal energy expenditure without becoming fatigued (Figure 2). Patients with mild SLE had little or no ability to meet the energy requirements for nonsedentary occupational and recreational tasks. The onset of fatigue, as well as exhaustion, were induced at lower levels of physical exertion in patients with mild SLE than in sedentary, but otherwise healthy, individuals of similar age, weight, and sex. At both ventilatory threshold and peak oxygen consumption, test duration was shorter, fewer test stages were completed, and workload was diminished in patients with mild SLE compared with sedentary controls. In women with mild SLE, the magnitude of aerobic insufficiency was great enough to not only limit the capacity to engage in occupational and recreational tasks but to even limit the ability to sustain normal activities of daily living for more than minimal periods of time. Before diminished treadmill exercise test performance can be determined to be due to low aerobic capacity, nonmetabolic factors (for example poor motivation and joint or muscle pain occurring during exercise) must be excluded as reasons for early test termination. Exclusion of these factors can be accomplished by using objective, physiologic criteria that correlate with maximal aerobic capacity, such as a peak respiratory exchange ratio of greater than 1.10 and a peak heart rate that is at least 90% of the age-predicted value. Patients in the current study attained peak respiratory exchange ratio and peak heart rate levels that exceeded these criteria, indicating that they approached maximum aerobic capacity at peak exercise and that exercise was not limited by nonmetabolic factors. Results of the current study indicate that maximal oxidative metabolism in patients with mild SLE is attenuated to a degree that has been previously reported only in patients with severe cardiovascular, pulmonary, and metabolic disease (21–26).

At any intensity, an individual's ability to sustain physical activity is determined by her or his capacity to consume oxygen to supply the necessary energy. The ability to utilize oxygen effectively depends on effective oxygenation of hemoglobin, delivery of the oxygenated hemoglobin to the target tissues (in this case the exercising muscles), uptake of oxygen from the blood into the cells of the tissue (muscle fibers), and utilization of the oxygen by the respiratory chain to supply adenosine triphosphate for energy. Disruption of the oxidative metabolic pathway at any point between the oxygenation of hemoglobin and the respiratory chain results in decreased energy availability and physiologic fatigue. In the current study, patients with mild SLE had no clinical evidence of cardiac or pulmonary disease so it was unlikely that aerobic capacity was attenuated by the limitations placed upon oxygen diffusion from the lung to the hemoglobin or impairment of delivery of oxygen to the tissues by central circulation. Muscle tissue analyses have revealed type-II selective atrophy (27–29), impaired excitation-contraction coupling (28, 29), and microcirculatory lesions (27, 29, 30) in biopsies taken from patients with SLE. It is reasonable to suspect that the most likely mediators of FAI in patients with SLE are microangiopathy leading to diminished oxygen delivery to the muscle cells, attenuated ability to extract and use oxygen in the muscle tissues, or a combination of these factors.

In conclusion, 14 of the 18 women with mild SLE had an FSS greater than 4.0, indicating that fatigue was perceived to limit the ability to participate in routine physical and social activities. Twelve of the 14 patients with an FSS greater than 4.0 had FAI. Sixteen of the 18 patients with mild SLE (89%) reported that fatigue limited their ability to engage in physical activity. FSS correlated with physiologic deficits in aerobic capacity and treadmill work capacity. The women with mild SLE had FAI, indicating that aerobic capacity was decreased to levels observed only in patients with debilitating diseases. These findings confirm that perceived activity-limiting fatigue is extremely common even in patients with mild SLE. A substantial portion of this fatigue appears to be due to severe impairment of the oxidative metabolic pathway. Impairment of oxidative metabolism was determined to limit these patients' ability to engage in physical activity. In the women with mild SLE, exhaustion during treadmill testing occurred at energy levels required for performing activities of daily living, leaving no reserve for sustained activities or more strenuous occupational and recreational activities. Based on objective, physiologic measures of fatigue and known energy requirements for activities of daily living, occupational tasks, and recreational activities, the results of this study have presented evidence of aerobic insufficiency as a cause of fatigue and disability in women with mild SLE.


The authors thank Dr. Marc Hochberg for his input and guidance in the initial stages of manuscript preparation and submission.