Exploratory Study on VO₂ On-kinetics During Treadmill Walking in Women With Systemic Lupus Erythematosus

Abstract

Objective

To determine if oxygen consumption (VO₂) on-kinetics differed between groups of women with systemic lupus erythematosus (SLE) and sedentary but otherwise healthy controls.

Design

Exploratory case control study.

Setting

Medical school exercise physiology laboratory.

Participants

Convenience samples of 12 women with SLE and 10 sedentary but otherwise healthy controls.

Intervention

None.

Main Outcome Measurements

VO₂ on-kinetics indices including time to steady state, rate constant, mean response time (MRT), transition constant, and oxygen deficit measured during bouts of treadmill walking at intensities of 3-METS and 5-METS.

Results

Time to steady state and oxygen deficit were increased and rate constant was decreased in the women with SLE compared to controls. At the 5-MET energy demand, the transition constant was lower and MRT was longer in the women with SLE than in the controls. For a comparable, relative energy expenditure that was slightly lower than the anaerobic threshold, the transition constant was higher in the controls than in the women with SLE.

Conclusion

VO₂ on-kinetics was prolonged in the women with SLE. The prolongation was concomitant with an increase in oxygen deficit and may underlie performance fatigability in women with SLE.

Systemic lupus erythematosus is a collagen vascular disease that affects as many as 1.5 million people in the U.S. alone¹. Ninety percent of patients with SLE are women and the disease is most prevalent in the second to fifth decades of life^{2, 3}. Fatigue is among the top three debilitating symptoms of SLE and persists in nearly all patients who have the disease^4–8. In patients who have SLE, the severity of fatigue^{9, 10} and physical disability⁹ have been strongly associated with decreases in aerobic capacity^9–13.

Phenotypes of aerobic capacity include peak VO₂, anaerobic threshold, and VO₂ on-kinetics¹⁴. Each phenotype characterizes a different cardiorespiratory response to physical activity and is regulated by a separate mechanism. For example, peak VO₂ is an index of the maximum rate at which the aerobic system can supply energy^{14, 15}. The rate-limiting factor for peak VO₂ is thought to be attainment of maximum cardiac output (Qt) but other factors can limit peak VO₂ as a result of pathomechanisms^{14, 15}. Severely low peak VO₂ has been reported in patients who had SLE compared to healthy controls^{9, 10, 13} and patients who had anemia or hypertension¹¹. Peak VO₂ has been strongly and indirectly associated with self-reports of fatigue severity^{9, 10} and physical disability⁹ in patients with SLE.

The second phenotype, anaerobic threshold, is the VO₂ corresponding to the onset of anaerobic by-product accumulation during cardiopulmonary exercise testing^{14, 16, 17}. A collage of factors, the confluence of which is not completely understood¹⁵, appears to mediate the anaerobic threshold. When aerobic energy output is insufficient for meeting the total energy expenditure, the deficit must be compensated by increased anaerobic glycolysis. High levels of glycolytic activity result in the accumulation of anaerobic by-products in plasma and muscle cells^{14, 15, 17}, which mechanizes work induced fatigue and physical activity intolerance¹⁸. Severe decreases in anaerobic threshold have been reported in patients with SLE^{9–11, 13}. In one study, women with SLE approached their anaerobic threshold while walking on a treadmill at energy demands of only 3-METS¹⁰. Healthy controls reached their anaerobic threshold at approximately 5-METS. For most healthy adults, energy expenditures of between 3-METS and 5-METS are of moderate intensity^{19, 20} and easily tolerated.

VO₂ on-kinetics reflects the rapidity with which the aerobic system can transition to a state of increased energy output^{14, 21, 22}. From the onset of physical work, total energy output must be sufficient for meeting the energy demand. However, aerobic energy output increases gradually over time (Figure 1), continuing until the energy output of the system reaches a steady state that meets the total demand^{14, 21, 23}. Until the steady state is attained, anaerobic glycolysis compensates for what would otherwise be the energy deficit (oxygen deficit). Because the anaerobic energy contribution is indirectly determined by the rate at which aerobic energy output increases, slowing of VO₂ on-kinetics may impair the ability to sustain physical activity. VO₂ on-kinetics appears to be regulated by microcirculatory dynamics^24–26, and mitochondrial function^27–29. Pathological adaptations occurring in the muscle capillary bed of patients with SLE could impair microvascular reactivity and oxygen diffusion^30–34. However, a report on VO₂ on- kinetics in patents with SLE was not found.

Figure 1.

Oxygen consumption during continuous exercise. t₀ is the onset of exercise and t₁ is the time taken to reach a steady state or an asymptote and (t₆) is 6-minute test endpoint. Area between Line A and the X-axis the resting VO₂ and Line B represents the energy demand of the activity being performed. The area above the trajectory line and under Line B is the oxygen deficit. VO₂ is the amplitude of change in oxygen consumption at the 6-minute test endpoint.

The purpose of this study was to provide an initial characterization of VO₂ on-kinetics during treadmill walking in women with SLE. VO₂ on-kinetics was examined at what are typically moderate intensities for most healthy adults in who cardiorespiratory function is uncompromised. The hypothesis that VO₂ on-kinetics may be diminished in women with SLE was tested. The main analysis compared five indices of VO₂ on-kinetics between a group of women with SLE and a group of sedentary but otherwise healthy controls.

METHODS

Subjects

Twelve sedentary women who had SLE and 10 sedentary but otherwise healthy women participated in this study (Table 1). Ages ranged from 27 to 57 years in the subjects with SLE and from 30 to 43 years in the controls. Subjects denied participating in any physical activity that caused them to perspire for 10 minutes or longer, once or more per week, in the six months prior to participation. Subjects with SLE met at least four of the 11 American College of Rheumatology 1982 Revised Criteria for the Classification of Systemic Lupus Erythematosus³⁵. The women with SLE had Systemic Lupus Activity Measure scores indicating their disease activity had been negligible to mild with no flares for the previous 30 days. Six of the women with SLE were on maintenance doses of prednisone (mean dose = 9.2 ± 3.8 mg/day, n=6); two were on five mg/day, three were on 10 mg/day, and one was on 15 mg/day. None of the women with SLE had comorbidities known to impair treadmill walking. Exclusionary conditions included circulatory, chronic or restrictive pulmonary, renal, neurological, or musculoskeletal diseases, which would adversely affect cardiorespiratory function. Those with severe anemia (plasma [HB] < 12 g/dl), fibromyalgia or chronic fatigue syndrome were also excluded. None of the subjects was taking medications known to limit or enhance exercise tolerance or aerobic capacity.

Table 1.

Demographic Characteristics and Maximum Treadmill Exercise Test Responses

	SLE	Control
Age (years)	38.3 ± 09 .4	39.1 ± 04.0
Height (cm)	160.80 ± 06.5	165.60 ± 08.0
Weight (kg)	70.03 ± 09.51	72.50 ± 15.73
Body Mass Index (kg/meter2)	27.42 ± 05.65	26.48 ± 05.44
Peak Respiratory Exchange Ratio	1.19 ± 0.07	1.29 ± 0.07
Attained Peak Heart Rate (beats/min)	163 ± 17	168 ± 29
Predicted Peak Heart Rate (beats/min)	182 ± 10	181 ± 04
% Predicted Max Heart Rate	90.2 ± 0.09	91.0 ± 0.17
Peak VO2 (ml/kg/min)	20.1 ± 04.7*	28.9 ± 05.1
Anaerobic Threshold (ml/kg/min)	12.9 ± 02.9**	17.9 ± 03.1

Data are means ± one standard deviation unit. BMI = body mass index, RER = respiratory exchange ratio,

^*Significantly different from controls; p=0.0001; 95% CI: 03.7 to 15.3 ml/kg/min,

^**Significantly different from controls; p=0.0001; 95% CI: 01.5 to 09.3 ml/kg/min

This protocol was reviewed and approved by the institutional, human subjects review board prior to its beginning. Informed consent was obtained from each subject prior to participation in accordance with institutional policies and the Declaration of Helsinki³⁶.

Apparatus and Formulae

Exercise tests were completed on a Trackmaster motorized treadmill^a and pulmonary gas exchange measurements were made using a Medgraphics Cardio 2, breath-by-breath cardiopulmonary exercise testing system^a. The system was calibrated before each test. Anaerobic threshold was determined by the V-slope method³⁷. Peak heart rate was determined electronically from an electrocardiogram by multiplying the six-second cardiac cycle rate by 10. Peak cardiac output (Qt) was measured by the exponential rise CO₂ rebreathing technique³⁸ applied at peak exercise. Peak arteriovenous oxygen difference (a-vO₂) was calculated by dividing peak VO₂ by peak Qt.

VO₂ on-kinetics was characterized using the monoexponential model developed by Whipp²³. VO₂ was plotted on time during continuous work rate tests (Figure 1). The algorithm, VO_2(t) = (ΔVO₂ + VO_2rest)(1−e^−tk) was then applied. In this model ΔVO₂ is the amplitude of increase in oxygen consumption at the sixth minute of exercise, t is time, and K is the rate constant. Attainment of a steady state was determined by the least squares method and confirmed by absence of a statistically significant increase in VO₂ between the third and sixth minutes of the exercise bout. Pearson product moment correlation coefficients (r) were calculated to quantify goodness of fit of the VO₂ kinetics algorithm. Goodness of fit was determined by the relationship between the iterations of VO₂ plotted on time generated by the algorithm and those actually measured. Significant correlation coefficients indicated high goodness of fit for the VO₂ on-kinetics model (Table 2). Eighty nine percent and 91% of the variances in VO₂ were attributed to the time domain in the women with SLE and controls respectively. Oxygen deficit was calculated as [6 minutes times steady state VO₂] minus the area below the VO₂-time curve (Figure 1). The overall VO₂ on-kinetics rate was quantified by a transition constant, calculated as the quotient of VO₂ and MRT ³⁹, where MRT is the mean response time.

Table 2.

Goodness of FIT for the VO₂ Kinetics Model

	Median	Mode	Mean	SD	Min	Max
Lupus	0.966	0.970	0.945	0.063	0.688	0.991
Controls	0.960	0.973	0.960	0.031	0.866	0.981

Pearson product moment correlation coefficients for relationships among VO₂-time iterations calculated from the model algorithm and those actually measured during the steady state tests. Data include both 3-MET and 5-MET energy demands for each group. SD is one standard deviation, Min is the lowest r-value, and Max is the highest r-value.

Procedure

Subjects first rested quietly in the supine position for at least 10 minutes. Subjects then completed a maximum treadmill test according to the modified Bruce protocol⁴⁰. The targeted stopping point for this test was the subject’s indication that she could not continue exercising despite strong encouragement from the testing staff. Pulmonary gas exchange and heart rate were measured throughout the test. Cardiac output was measured at peak exercise.

Forty eight to 164 hours after completing the maximum treadmill tests, subjects completed two randomly ordered, continuous work rate tests. A simple Latin square randomization method ensured equal ordering. These tests consisted of six minutes of sustained treadmill walking at 3-MET and 5-MET energy demands. The intensity of 3-METs corresponded to treadmill walking at a speed of 2.0 mph and an inclination of 1.5% grade^{19, 20}. The intensity of 5-METs corresponded to a speed of 2.0 mph and an inclination of 9.0% grade^{19, 20}. Subjects rested quietly for 10 minutes prior to the first test in the order. A recovery period separated the first test in the order from the second test. During recovery, subjects rested in the supine position for 20 minutes or longer, until heart rate and VO₂ returned to resting levels. VO₂ was measured continuously throughout these work bouts and VO₂ on-kinetics was determined from these measurements.

Statistics

Data were analyzed using the SAS, version 9.1, statistical analysis system^b. One and two-way ANOVA and ANCOVA were used for testing for intergroup, test, and interaction differences. The least squares means procedure was applied following the finding of a significant interaction in the two-way analyses. The general assumptions underlying these procedures are that ratio or interval data is being analyzed and that additional observations beyond our small sample (N=22) would trend toward a normal distribution of the mean according to the Central Limit Theorem⁴¹. Dependent variables measured during the maximum treadmill test were secondary outcome measures and included were peak VO₂, anaerobic threshold, Qt, a-vO₂, respiratory exchange ratio and heart rate. These data were assessed for significant group differences by one-way statistical analyses. The main outcome variables for this study were the indices of VO₂ on-kinetics including transition constant, time to steady state, rate constant, MRT, and oxygen deficit. These data were acquired during the submaximal, continuous work rate tests and assessed for significant group and test differences by two-way analyses. Statistical significance was set at p≤0.05. Reported means are adjusted for covariance when appropriate. Otherwise, data are reported as simple means ± one standard deviation.

RESULTS

Despite the difference in the age ranges, mean age and other group demographics were similar for the women with SLE and controls (Table 1). Gas exchange data obtained from the maximum treadmill tests are also provided in Table 1. Both groups attained a peak RER of greater than 1.15 and a peak heart rate of at least 90% of the age predicted peak heart rate, where predicted peak heart rate=220 bpm − 1 bpm per year of age. These criteria are accepted indicators that VO₂ approached a physiologically maximal level at volitional exhaustion^{19, 20, 42} and that tests were not stopped early due to motivational factors.

Age was not a significant covariate in the group comparisons of peak VO2, anaerobic threshold, Qt, a-vO₂, RER, or any of the measures of VO₂ on-kinetics. ANOVA indicated that peak VO₂ and anaerobic threshold were 30.4% and 27.9% lower in the women with SLE than in the controls, respectively (Table 1). Measurements of Qt were available in eight patients with SLE (14.0±0.31L/min) and eight controls (13.8±02.8 L/min). Qt was not significantly different between these subgroups. However, peak a-vO₂ was 21% lower (p<0.029 CI: 11.3 to 30%) in the women with SLE (10.5±02.4 vol%) than in the controls (13.3±02.1 vol%).

A significant difference in ΔVO₂ was not observed between the women with SLE and the controls (Table 3). ANOVA indicated that ΔVO₂ was significantly higher (p<0.002, 95% CI: 180 to 252 ml/min) for the 5-MET energy demand than for the 3-MET energy demand. A significant difference in ΔVO₂ was not observed between the third and sixth minutes of exercise in either group or between the 3-MET or 5-MET tests. ΔVO₂ covaried significantly with rate constant (p<0.002), MRT (p<0.015) and oxygen deficit (p<0.001). ΔVO₂ did not covary with transition constant, or time to steady state.

Table 3.

Changes in ΔVO₂ for the 3-MET and 5-MET Energy Demands

	3-MET		5-MET
	Lupus	Control	Lupus	Control
ΔVO2 (ml/min)*	599 ± 293	540 ± 144	766 ± 261	817 ± 225
ΔVO2(6–3) (ml/min)	−06.3 ± 66.7	−03.8 ± 56.4	05.8± 80.0	−05.6 ± 58.2

Data are means ± one standard deviation. ΔVO₂ is the change in oxygen consumption above rest at minute six of exercise, and ΔVO_2(6–3) is the difference in VO₂ above rest between the third and sixth minutes of exercise.

^*Significantly higher for 5-METS than for 3-METS (statistical main effect, p=0.0016, 95% CI: 180 to 252 ml/min)

Time to steady state, rate constant, and oxygen deficit responses are depicted in Figure 2. ANOVA indicated that time to steady state was significantly longer (p=0.0027; CI: 0.13 to 2.11 min) for the 5-MET test than for the 3-MET test. ANCOVA and ANOVA main effects indicated that rate constant was lower (p<0.005; 95% CI: 0.00 to 0.42 min⁻¹), time to steady state longer (p<0.016; 95% CI: −0.08 to 1.86 min), and oxygen deficit higher (p<0.028; 95% CI: −32 to 216 ml) in the women with SLE than in the controls.

Figure 2.

Time to steady state, rate constant, and oxygen deficit in women with SLE and controls. Error bars equal one standard deviation unit. * Significantly different main effects for women with SLE than for controls (p<0.005 rate constant, p<0.016 time to steady state, p<0.028 oxygen deficit)

Figure 3 describes the group and test differences in transition constant and MRT. ANOVA indicated that for the 5-MET test, the transition constant was lower (p<0.004; 95% CI: 0.393 to 7.647 ml/min/sec) and the mean response time longer (p<0.0170; 95% CI: −2.701 to 34.720 sec) in the women with SLE than in the controls(Figure 3). In the women with SLE, the transition constant did not increase from the 3-MET test to the 5-MET test. However ANCOVA indicated that MRT was significantly (p<0.0237; 95% CI: −02 to 32 sec) longer for the 5-MET test than for the 3-MET test in the women with SLE. In controls the transition constant significantly (p<0.0136; 95% CI: 0.094 to 6.08 ml/min/sec) increased from the 3-MET test to the 5-MET test, whereas test differences in their MRT were not observed.

Figure 3.

Transition constant and mean response time in women with SLE. Error bars are one standard deviation unit. * significantly higher than 3-METS (p<0.004), + significantly higher than 3-METS in the women with SLE, **significantly higher than 3 METS (p<0.017).

In the women with SLE, VO₂ at six minutes of walking at 3-METS was 88.5 ± 25.9% of the anaerobic threshold value. In the controls, VO₂ at six minutes of walking at 5-METS was 87.3 ± 23.8% of the anaerobic threshold value. Although these percentages of anaerobic threshold VO₂ were nearly identical between the groups, the transition constant was significantly (p<0.014; 95% CI: −0.20 to 7.02 ml/min/sec) higher in the controls than in the women with SLE (Figure 3).

DISCUSSION

In the women with SLE, VO₂ on-kinetics was prolonged in response to what are typically moderate and easily tolerated walking intensities for most healthy individuals. The maximum rate for VO₂ on-kinetics was attained at the 3-MET energy demand in the women with SLE, whereas a significant increase in the rate occurred between the 3-MET and 5-MET energy demands in the controls. Moreover, the VO₂ on-kinetics rate was lower in the women with SLE than in the controls, even at comparable levels of energy expenditure relative to and slightly below the groups’ anaerobic thresholds. Results of this study were in agreement with the hypothesis that VO₂ on-kinetics may be reduced in women with SLE, particularly during treadmill walking at work rates that are of moderate intensity for most healthy subjects.

In the control group, peak VO₂ was similar to age predicted, population normative values as determined by the well-established algorithm of Bruce^{43, 44}. Anaerobic threshold occurred on average at approximately 62% of peak VO₂, a level typical of healthy adults^{45, 46}. In addition, the MRT was similar to previous reports in healthy, sedentary adults^47–49. These findings suggest that, in the controls, the aerobic response was indicative of those expected for sedentary but otherwise healthy women in the general population.

The magnitude of the observed decreases in peak VO₂, anaerobic threshold, and VO₂ on-kinetics in the women with SLE in this study were similar to those previously reported in patients with chronic illnesses such as congestive heart failure^{50, 51}, chronic obstructive pulmonary disease⁵² and highly active antiretroviral therapy for HIV infection^{39, 53, 54}. Peak VO₂ in the women with SLE was also far below values reported after 10 days of bed rest in healthy women of similar ages⁵⁵. In the women with SLE, peak VO₂ was 32.0 ± 13.3% lower than levels expected for healthy, sedentary women of similar ages, as determined by the well established algorithm of Bruce^{43, 44}. The decreases in peak VO₂^{9, 11, 13, 56}, and anaerobic threshold^{11, 13, 53, 56} in the women with SLE in this study were also similar to previous reports on patients with SLE. All combined, the number of observations in these reports was substantial. No previous reports were found in which peak VO₂ or anaerobic threshold in patients with SLE were similar to or higher than healthy controls’. In concert, the totality of observations suggest that the trends observed in this and the previous studies may reflect activity-induced, metabolic responses that are characteristic of the population of patients with SLE in general. However, previous information is lacking regarding VO₂ on-kinetics in patients with SLE.

The initial component of VO₂ on-kinetics is a fast response in which VO₂ increases rapidly towards the steady state. Proposed regulators⁵⁷ of the fast component include central circulatory oxygen delivery^58–60, capillary blood flow dynamics^24–26, and dynamics of muscle fiber recruitment^{24, 61}. Mitochondrial inertia and respiratory chain inhibition have also been suggested as fast component regulators^{27–29, 27, 28, 62}. Group similarity in peak Qt suggested that central circulatory oxygen delivery was not impaired in these women with SLE. Conversely, the large reduction in peak a-vO₂ suggested that aerobic metabolism could have been limited by restricted muscle oxygen extraction. Microvascular adaptations have been observed in skeletal muscle samples obtained from individuals who had SLE^30–34. These adaptations could impair muscle oxygen extraction and slow the fast component by diminishing the microvascular reserve and impeding oxygen influx diffusion.

A plateau in the VO₂-time plot follows the fast component, indicating that an aerobic steady state has been achieved with the total energy demand^{14, 21, 23}. However, the plateau is often replaced by a slower rise of VO₂ that delays the onset of the steady state when the work being performed is of severe intensity⁶³. This slow component may be less likely to occur during treadmill exercise than during cycling⁶⁴. Some^{29, 65–67} but not all⁶⁸ studies have suggested that progressive recruitment of Type-II muscle fibers may mediate the slow component. Predominance of Type-I muscle fibers and selective Type-II muscle fiber atrophy have been observed in muscle samples obtained from individuals with SLE⁶⁹. This adaptation is not completely understood but could possibly attenuate a slow component transition.

VO₂ on-kinetics is slower in untrained individuals than in those who are more physically fit^{70, 71}. Aerobic exercise training may improve VO₂ on-kinetics^72–74. While aerobic exercise training has been reported to increase peak VO₂ and anaerobic threshold in patients with SLE^{12, 75}, further research is needed to determine whether exercise training can improve VO₂ on-kinetics in patients with SLE.

Study Limitations

Half of the women with SLE in the current study were maintained on non-fluorinated, orally administered prednisone with doses well below 40 mg/day. Drug induced myopathy is one possible side effect of this medication, particularly when the agent is fluorinated or dosages exceed 40 mg/day⁷⁶. Orally administered prednisone dosages, similar to those used by patients in the current study, have been shown to neither decrease mitochondrial enzyme concentrations⁷⁷ nor reduce peak VO₂⁷⁸. However, the effect of prednisone on VO₂ on-kinetics is yet unknown.

Physically inactive lifestyle was an inclusion criterion of this study. However, routine physical inactivity was determined subjectively and by a single question. Because patients with SLE may be more fatigable, physical activity levels could have been slightly lower in the women with SLE than in the controls. Bostrom and associates reported that peak VO₂ was not related to daily physical activity levels in women with SLE⁵⁶. However, it remains unknown as to whether VO₂ on-kinetics is associated with daily physical activity in this patient population.

The rate of VO₂ on-kinetics may be increased by prior, strenuous exercise^79–81. Equal randomization of test order and a return of VO₂ to baseline levels between the tests were used in an attempt to minimize this effect. An equal number of subjects completed each of the testing orders in both of the groups. However, the 5-MET energy demand was above the anaerobic threshold only in the women with SLE so it may have been more strenuous for them than for controls. Bias in the distribution of this effect may have been introduced despite equality in the test order. If so, the rate of rise in VO₂ during the 3-MET exercise bout could have been slightly increased and the time taken to achieve the steady state slightly decreased, by prior exercise, in the women with SLE who completed the 5-MET test before completing the 3-MET test.

Menstrual cycle phase and use of oral contraceptives were not controlled in this study. Dean and associates reported that in healthy women, menstrual phase had no effect on peak VO₂ or lactate threshold⁸². Lebrun et al⁸³ reported a small decrease in peak VO₂ of 4.7% from the follicular to the mid-luteal phase in a group of healthy women using oral contraceptives. However, a slight increase of 1.4% occurred across these phases in a placebo control group. Similar effects would account for a variance of no greater than -20% to +06% of one standard deviation from the mean peak VO₂ in the women with SLE (−0.95 to +0.28 ml/kg/min) and −27% to +08% in the controls (−1.36 to +0.40 ml/kg/min). Thus the potential bias introduced by variance in the menstrual phase and uncontrolled use of oral contraceptives appears to be minimal.

For this study, a convenience sample of patients with SLE was recruited in which volunteers were identified and selected from patient records. Due to the non-randomized selection of the small sample of subjects, selection bias could have affected the interpretation of the results. Therefore, even though there were response similarities among these subjects with SLE and numerous other reports, generalizations to the overall population of subjects with SLE must be made with due caution.

CONCLUSIONS

VO₂ on-kinetics during treadmill walking was slower in the women with SLE than in sedentary but otherwise healthy controls. Oxygen deficit was higher in the women with SLE than the controls. It is possible that impaired VO₂ on-kinetics may contribute to performance fatigability in patients with SLE.

List of Abbreviations

ANCOVA

analysis of covariance

ANOVA

analysis of variance

SLE

Systemic lupus erythematosus

METS

metabolic equivalents

MRT

mean response time

RER

respiratory exchange ratio

VO₂

oxygen consumption