VO₂ On-kinetics and Aerobic Exercise Training in Females with Systemic Lupus Erythematosus

Abstract

Fatigue among females with systemic lupus erythematosus (SLE) is commonly reported as the most debilitating symptom. Slower oxygen uptake (VO₂) on-kinetics have been observed among females with SLE, when compared to healthy control individuals. Aerobic exercise training may improve VO₂ on-kinetics, however, this has not been examined in the SLE population.

Objective:

To characterize VO₂ on-kinetics in SLE and examine changes after 12 weeks of aerobic exercise training (AET).

Methods:

Females (n=16; 42 ± 10 years) with SLE (SELENA-SLEDAI ≤ 4) completed a cardiopulmonary exercise test (CPET) and a constant work rate test (CWRT), before and after participation in supervised vigorous treadmill AET for 12 weeks. Pulmonary gas exchange, heart rate (HR) and deoxygenation ([HHb]) of the right gastrocnemius muscle were also measured during the CWRT. Model fits for VO₂, HR and [HHb] were performed, and a ratio of the increase in [HHb] and VO₂ were examined in a subset of participants (n=10). On-kinetic responses of all variables and the [HHb]/VO₂ ratio were analyzed before and after AET using paired sample t-tests or Wilcoxon signed-rank tests. Spearman’s rank correlations were used to examine the relationships of VO₂ on-kinetics and peak CPET variables.

Results:

VO₂ on-kinetics was faster (mean difference [95% CI], p-value: −12.0 s [−18.1 to −6.0], p <0.001) following AET, and the VO₂ time constant post-AET was moderate and indirectly related to cardiorespiratory fitness outcomes for peak VO₂ (r_s = −0.562, p = 0.024) and peak work rate (r_s = −0.535, p = 0.033). The HR on-kinetics was also faster following AET (median difference [95% CI], p-value: −6.7 s [−30.8 to −1.2], p = 0.022), however a change in the [HHb] time course was unlikely (median difference [95% CI], p-value: −1.1 s [−3.6 to 23.5], p =0.508).

Conclusion:

Following supervised exercise training, females with SLE demonstrated faster VO₂ on-kinetics, which were also related to cardiorespiratory fitness. The faster VO₂ on-kinetics with exercise training was also accompanied by faster HR on-kinetics, suggesting central circulatory adaptation occurring in these females with SLE.

Background

Systemic lupus erythematosus (SLE) is a complex autoimmune disease affecting a higher proportion of females than males, and individuals in early to middle adulthood.¹ Between 67% and 90% of patients with SLE report symptoms of fatigue^{2, 3} with many stating that fatigue is their single most frequent and debilitating symptom.⁴ SLE-related fatigue remains a challenge that has important implications for the quality of life of patients throughout their lifespan.

Performance fatigability is the inability to sustain a given intensity, duration or frequency of physical effort, and informs regarding the susceptibility of an individual to fatigue limitations during sustained physical activity.⁵ It can be objectively measured during standardized tests, to include VO₂ on-kinetics during a constant work rate test (CWRT).⁶ Pulmonary O₂ uptake (VO₂) represents the body’s overall O₂ consumption, which is a precise interaction by the collective respiratory, circulatory and metabolic system pathways.⁷ At the onset of an increase in work rate, VO₂ rises gradually until it attains an oxidative steady state with the oxygen demand. The time required to attain the oxidative steady state (i.e., VO₂ on-kinetics) reflects dependency on anaerobic metabolic supplementation, which is necessary to meet the pre-steady state, immediate energy requirements.⁸ Untoward by-products of anaerobic supplementation contribute to muscular fatigue,⁹ thus an individual with slower VO₂ on-kinetics is considered more fatigable compared to one with a faster oxidative response.

Slower VO₂ on-kinetics has been reported among less fit but otherwise healthy individuals^{10, 11} and in those with chronic illnesses,^{12, 13} particularly females with SLE.¹⁴ Field tests have suggested that performance fatigability was reduced following aerobic exercise training programs in those with SLE,^{15, 16} however information regarding potential changes in VO₂ on-kinetics have not been investigated.

This pilot study aimed to examine the influence of 12 weeks of aerobic exercise training (AET) on VO₂ on-kinetics in females with SLE. It was hypothesized that after AET, faster VO₂ on-kinetics would be observed during sub-maximal exercise.

Methods

This pilot study was conducted to provide preliminary data for future and larger clinical trials. It was stopped early due to the COVID-19 pandemic and restrictions to the study site in 2020. This pilot study was approved by the Institutional Review Board at the National Institutes of Health (NIH) and registered on clinicaltrials.gov (NCT03186794). Written consent was obtained from all participants prior to formal enrollment and study participation. Results from this study report on the same participants as those in a previous publication.¹⁷

Participants

No formal sample size calculations were performed for this pilot study, which aimed to enroll up to 20 evaluable participants. Females aged between 21 and 80 years old and living within the greater Washington D.C. area were recruited from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) SLE Natural History and Pathogenesis protocol. The assessments and exercise training sessions were all conducted at the NIH Clinical Center in Bethesda, Maryland.

Participants were considered sedentary, reporting significant fatigue (Fatigue Severity Scale (FSS) score ≥ 3) and had stable SLE disease activity that was mild or less (SELENA-SLEDAI ≤ 4). Each participant also met the American College of Rheumatology Criteria for the Classification of SLE¹⁸ and were stable on immunosuppressive medications, prednisone or steroid equivalent in the previous 3 months. In addition, participants were excluded for any of the following: inability to walk on a treadmill; BMI ≥ 40 kg/m²; taking ≥ 15 mg daily of prednisone (or equivalent); had infusion of Rituximab within the previous 6 months or started methotrexate, mycophenolate mofetil, azathioprine, cyclophosphamide or biologics within the previous 3 months; presence of primary/secondary medical conditions (e.g., cardiomyopathy, pulmonary vascular disease) or taking medications that may limit exercise capacity, make it unsafe to conduct exercise or adapt to exercise training; present symptoms of cardiopulmonary conditions; significant cardiac, pulmonary, hepatic, or renal dysfunction; chronic conditions including diabetes mellitus, fibromyalgia; anticoagulation, anemia, severe Raynaud’s phenomenon, significant peripheral vascular disease, uncontrolled or untreated thyroid dysfunction; previous history of cerebrovascular accident; history or presence of most cancers; deep vein thrombosis; HIV infection; current smoker or active substance abuse; pregnancy or nursing; orthopedic conditions limiting ability to walk on the treadmill; inability to give informed consent.

Study Design

All eligible participants completed a treadmill cardiopulmonary exercise test (CPET) for evaluation of cardiorespiratory fitness and ventilatory anaerobic threshold, as well as a sub-maximal constant work rate test (CWRT) for determination of VO₂ on-kinetics on a separate day. These tests were repeated after 12 weeks of supervised aerobic exercise training (AET). Testing was mostly performed in the morning, with tests repeated at approximately the same time of day for each participant and conducted in the order as specified below. The CPET and CWRT was performed in a laboratory that maintained a consistent temperature, and environmental conditions (i.e., ambient temperature, relative humidity and barometric pressure) were recorded by the gas exchange system during calibrations at the time of the tests. Participants were given standard instructions in preparation for testing, to include wearing appropriate clothing and footwear for walking on a treadmill, as well as avoiding strenuous activity and to refrain from caffeine products prior to testing.

Study Procedures

Cardiopulmonary Exercise Test (CPET):

Participants walked on a treadmill to a ramp increase in work rate while breath-by-breath pulmonary gas exchange was measured (CardiO₂ Ultima; MGC Diagnostics Corp, St. Paul, MN, USA). The increase in speed, followed by grade of the treadmill was individualized to the walking speed that was safe for participants to maintain walking uphill, and the same parameters were used in follow-up testing. The ventilatory anaerobic threshold determined by the gas exchange system was verified offline independently by two reviewers using the V-slope and ventilatory equivalent methods⁷ of the data displayed as averages of 8 breaths. Peak VO₂ was determined as an average of the last 20s achieved by the participant, prior to test termination due to volitional exhaustion. Time to peak exercise, peak HR and peak work rate achieved on the CPET were also variables of interest related to cardiorespiratory fitness.

Constant Work Rate Exercise Test (CWRT):

Participants performed sub-maximal treadmill walking consisting of 3 repeats of continuous moderate exercise for 6 minutes, interspersed with low intensity walking for 8 minutes. Pulmonary gas exchange was measured throughout, together with heart rate (HR) by 3-lead electrocardiogram (Enduro PhysioFlow, Manatec Biomedical, France) and concentration changes of muscle deoxygenation ([HHb]) of the right gastrocnemius by near-infrared spectroscopy (NIRS; PortaMon, Artinis Medical Systems, Netherlands). The initial 3-minute baseline walking and 8-minute light walking period was conducted at 0.7 mph without any incline. Moderate exercise was an increase in treadmill speed and/or grade that corresponded to ~85% of the ventilatory anaerobic threshold measured during the initial CPET. After exercise training, CWRT was repeated at the same treadmill speed and grade as the initial CPET (i.e., absolute work rate). A second CWRT was also performed using the updated treadmill speed and grade associated with ~85% of the ventilatory anaerobic threshold measured from the post-AET CPET (i.e., relative work rate), and these results are provided as a Supplement (Supplement Table 1).

Exercise Training:

All participants completed supervised AET of 12 weeks in duration, consisting of treadmill walking 3 times a week for 30 minutes at a target training HR range of 70 to 80% of heart rate reserve (HRR: Equation 1) determined from the baseline CPET.

$$\mathrm{H}\mathrm{R}\mathrm{R}=0.70\mathrm{a}\mathrm{n}\mathrm{d}0.80(\mathrm{P}\mathrm{e}\mathrm{a}\mathrm{k}\mathrm{H}\mathrm{R}-\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{H}\mathrm{R})+\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{H}\mathrm{R}$$ (Eq. 1)

During the training session, HR, treadmill speed/grade, and rating of perceived exertion (RPE) were recorded. Heart rate monitors (H7, Polar Electro Inc., Bethpage, NY) were worn during each session to help ensure participants exercised in the desired HR range. Each session included approximately 5 minutes of warm-up and cool-down, as appropriate. Participants were required to complete at least 80% of the exercise sessions (29 of 36) to be included in the study.

Data Analysis

VO₂ on-kinetics:

Examination of VO₂ on-kinetics from the CWRT was performed as previously described.⁶ Briefly, raw breath-by-breath data were filtered to remove data unrelated to the exercise response. Each individual exercise bout was then time-aligned, interpolated and averaged to create a single profile for each participant representing before AET (Pre) and after AET (Post). The VO₂ on-kinetics fit was started at the phase 1/phase 2 transition and modeled (OriginPro 2023, OriginLab Corp, Northampton, MA) using the mono-exponential function (Equation 2):

$${\mathrm{V}\mathrm{O}}_{2\left(\mathrm{t}\right)}={\mathrm{V}\mathrm{O}}_{2\text{(baseline)}}+\mathrm{\Delta }{\mathrm{V}\mathrm{O}}_2\left[1-{\mathrm{e}}^{-(\mathrm{t}-\mathrm{T}\mathrm{D})/\tau }\right]$$ (Eq. 2)

where VO_2(baseline) is the 60s mean prior to the increase in work rate; ΔVO₂ is the amplitude of the response; $\mathrm{\tau }$ and TD is the respective time constant and time delay of the model response. The fit was made to 4 minutes of the exercise bout, which optimized the fitting window without biasing the response.¹⁹

HR and Muscle deoxygenation ([HHb]) on-kinetics:

To further examine the components contributing to the VO₂ on-kinetic response before and after AET, HR on-kinetics representing central O₂ delivery and [HHb] on-kinetics representing local muscle O₂ utilization²⁰ were analyzed in a subset of these participants. The HR response for one participant could not be fit with the mono-exponential equation, and an attenuated [HHb] response was observed in 5 participants with high adipose tissue thickness at the NIRS measurement site. Therefore, this sub-analysis included the remaining 10 participants.

Second-by-second HR data were filtered and analyzed in a similar manner to the VO₂ on-kinetics section above, with the fitting window being the first increasing point after the start of exercise to 4 minutes of the exercise bout.

The second-by-second [HHb] signal was time-aligned and averaged to create a single response for each participant as the VO₂ on-kinetics section above, however, the fitting window was constraint to begin after the “observed” time delay to 90s of the exercise bout.¹⁹ The “observed” time delay was an average of the time associated with the first consistently increasing [HHb] point, assessed from each single exercise bout. Summed with the time constant from the [HHb] model fit, these are reported as the effective [HHb] time course.¹⁹

To describe the proportional rate of increase for local muscle O₂ extraction and pulmonary O₂ uptake at exercise onset, calculation of normalized [HHb] and VO₂ was performed as described by Murias et al.¹⁹ Briefly, normalization of the second-by-second [HHb] and interpolated VO₂ data were made to coincide with 0% (representing the baseline prior to exercise) and 100% (representing the exercise bout steady state) for each participant before AET (Pre) and after AET (Post). The ratio was determined by dividing the normalized [HHb] response with the normalized VO₂ response that was left-shifted by 20s to coincide with the approximate start of the phase 2 VO₂ on-kinetic response.¹⁹ A ratio of 1.0 indicates perfect matching between muscle O₂ extraction and utilization. In addition, the mean ratio between 20s and 130s was calculated to reflect the approximate start to when steady state was considered to have been reached in all participants.

Statistical Analysis

Difference in the on-kinetic responses of VO₂, HR, [HHb] and [HHb]/VO₂ ratio (i.e., after AET – before AET) was assessed for normality using the Shapiro-Wilk test, and paired sample t-tests or Wilcoxon signed-rank tests were used for parametric and non-parametric distributions, respectively. Effect sizes (Hedge’s g or correlation coefficient r) were considered small (g = 0.20; r = 0.10), moderate (g = 0.50; r = 0.30) or large (g = 0.80; r = 0.50). Due to resistance to outliers, Spearman’s rank correlations were used to examine the relationships of the time course of VO₂ on-kinetics and CPET variables. Size of the correlation coefficient (r_s) was considered as: 0.0–0.30 (negligible), 0.30–0.50 (low), 0.50–0.70 (moderate), 0.70 to 0.90 (high) and 0.90–1.00 (very high). All analyses were performed using SPSS 29.0.2.0 (IBM Corporation, Armonk, NY, USA). P-values are based on two-tailed analyses and reported as continuous values (instead of dichotomized). Data are reported as mean (± SD), unless otherwise indicated. The 95% confidence intervals (CI) for mean and median differences are reported as [lower to upper] bounds. The 95% CI for median difference used the Hodges-Lehmann estimator for Wilcoxon signed rank tests, while 95% CI for Spearman’s correlations was reported using the Caruso and Cliff’s method. All results were interpreted in the context of compatibility, magnitude of effect and observed range of variability along with p-values.

Results

Sixteen females with SLE completed 12 weeks of AET. A full description of the participants from this study can be found in a previously published paper.¹⁷ On average, participants were 42 (10) years of age with a BMI of 28.8 (5.5) kg/m². The average disease duration was 9.1 (6.5) years, with a SELENA-SLEDAI score of 1.4 (1.9). No changes were made to medications throughout the study, including prednisone dosages, which were 5mg on average. Participants’ disease activity at the time of enrolment was low, and no flares or changes in disease status were observed during the study.

Attendance was high at 92.2 (7.3) % with participants completing an average of 33 sessions out of a possible 36 sessions (range: 29 to 36 sessions). Total exercise time over the 12-week intervention averaged 996 (79) minutes (range: 870 to 1080 minutes). Across the training sessions, participants reached an average HR of 135 (13) bpm (range: 114 to 155 bpm) and HRR of 74 (3) %. Participants spent an average of 95 (4) % of their exercise session in or above the desired HRR range.

Peak cardiorespiratory fitness outcomes were presented previously, with improvements observed after AET for peak VO₂, work rate and exercise time.¹⁷ Parameters of the VO₂ on-kinetics fit for all 16 participants are shown in Table 1, with Spearman’s correlation outcomes for the VO₂ time constant (τ) and CPET variables shown in Supplement Table 2. The τVO₂ was smaller after AET (t(15) = −4.23, p < 0.001, g = −1.003), with a large effect. Average pre-training τVO₂ for participants was 33 (17) s and became faster to 21 (9) s following AET (Figure 1A). The change in τVO₂ in response to AET was very high and inversely related to the initial τVO₂ (r_s(16) = −0.926 [−0.975 to −0.791], p < 0.001) (Figure 1B). The τVO₂ was also observed to be moderate and inversely related to CPET outcomes of physical fitness for peak VO₂ and peak work rate (Figure 2) after AET (peak VO₂: r_s(16) = −0.562 [−0.832 to −0.076], p = 0.024; peak work rate: r_s(16) = −0.535 [−0.820 to −0.038], p = 0.033).

Table 1.

VO₂ On-Kinetic Fitting Parameters (n=16)

	Pre	Post	Mean Diff [95% CI]	p-value
Baseline (ml/min)	409 (120)	390 (95)	−19 [−55 to 17]	0.272
Amplitude (ml/min)	292 (80)	279 (79)	−13 [−39 to 14]	0.313
Time Delay (s)	14.1 (8.8)	16.3 (7.3)	2.2 [−4.0 to 8.4]	0.461
Time Constant (s)	32.8 (16.6)	20.8 (9.2)	−12.0 [−18.1 to -6.0]	<0.001

Abbreviations: Pre, pre-exercise training; Post, post-exercise training; Diff, difference; CI, confidence interval.

Figure 1:

Individual τVO₂ on-kinetics with mean (±SD) shown before (white, Pre) and after (black, Post) 12 weeks of AET. P-value between Pre and Post shown. B.) The relationship between initial τVO₂ on-kinetics at baseline and magnitude of change after 12 weeks of AET.

Figure 2:

The relationship between τVO₂ on-kinetics and indices of cardiorespiratory fitness from the CPET (peak VO₂, A; and peak work rate, B). Individuals with line of best fit shown before (white circles, grey line) and after (black circles, black line) 12 weeks of AET.

To further examine the components contributing to changes in τVO₂ on-kinetics, AET-induced adaptations on central O₂ delivery and local microvascular O₂ extraction were made by examining the HR on-kinetics and effective [HHb] time course, respectively, during the CWRT. Parameter fits for VO₂, HR and [HHb] were compared and shown in Table 2.

Table 2.

On-Kinetic Fitting Parameters for VO₂, HR and [HHb] (n=10)

	Pre	Post	Mean or Mediana Diff [95% CI]	p-value
VO2 Parameters
Baseline (ml/min)	342 (84)	342 (65)	0 [−43 to 43]	0.998
Amplitude (ml/min)	285 (90)	268 (73)	−16 [−48 to 16]	0.279
Time Delay (s)	12.4 (7.8)	14.6 (7.7)	2.2 [−7.1 to 11.6]	0.600
Time Constant (s)	34.8 (20.6)	23.2 (10.5)	−11.7 [−21.4 to −1.9]	0.024
HR Parameters
Baseline (beats/min)	87.2 (7.2)	81.7 (10.3)	−5.5 [−10.9 to −0.1]	0.046
Amplitude (beats/min)	15.1 (3.6)	13.8 (4.1)	−1.3 [−3.5 to 0.9]	0.216
Time Delay (s)	11.4 (9.2)	11.0 (7.3)	−1.7 [−3.4 to 2.5]a	0.959
Time Constant (s)	38.5 (26.3)	24.0 (7.5)	−6.7 [−30.9 to −1.2]a	0.022
[HHb] Parameters
Baseline (a.u.)	−6.8 (4.9)	−6.3 (1.8)	0.5 [−3.0 to 4.0]	0.738
Amplitude (a.u.)	2.3 (1.7)	2.5 (1.9)	0.2 [−0.6 to 1.1]	0.536
Time Constant (s)	9.4 (2.4)	13.2 (14.3)	−0.1 [−2.7 to 21.0]a	0.959
Average Time Delay (s)	8.2 (2.2)	8.2 (3.4)	0.1 [−2.2 to 2.3]	0.949
Effective Time Delay (s)	17.5 (3.5)	21.4 (16.6)	−1.1 [−3.6 to 23.5]a	0.508

Abbreviations: Pre, pre-exercise training; Post, post-exercise training; a.u., arbitrary units; Diff, difference; CI, confidence interval.

^aMedian difference and 95% CI by Hodges-Lehman for Wilcoxon Signed Rank test.

The HR on-kinetics were faster with a large effect following AET (Z = −2.293, p = 0.022, r = 0.725), meanwhile the effect of AET on the time course of [HHb] was minimal (Z = −0.663, p = 0.508, r = 0.210) (Table 2). Comparisons between these variables before and after AET are provided in Supplement Figure 1, with the change in τVO₂ and τHR on-kinetics with AET shown in Supplement Figure 2.

An “overshoot” in the [HHb] signal was observed at baseline, which remained after AET (Figure 3, inset), such that differences in the ratio for normalized [HHb] and normalized VO₂ was small following AET (median difference: 0.03 [−1.10 to 0.13], Z = 0.459, p = 0.646, r = 0.145).

Figure 3:

The time course of normalized [HHb] (square) and normalized VO₂ (circles) shown before (white symbols, A) and after (black symbols, B) AET, for the first 3 minutes of the CWRT. Figure 3 (inset): The ratio for normalized [HHb] and normalized VO₂ are shown across the same 3 minutes of the CWRT. A ratio value of 1 indicates perfect matching between [HHb] and VO₂. A profile of the ratio from before AET is shown in B as a black line for comparison purposes. An “overshoot” occurs in the first 60s, which is diminished but not completely abolished after AET.

Discussion

This pilot study expands our previous knowledge by characterizing changes in VO₂ on-kinetics following AET in females with SLE. Previous work has elucidated the presence of functional aerobic impairment²¹ and restricted VO₂ on-kinetics¹⁴ in females with mild to moderate SLE. The current study provides additional information regarding the potential influence of AET on physiological measures of performance fatigability in females with SLE.

Cardiorespiratory fitness reflects integration of the major systems to support engagement in prolonged or repetitious physical activity. Keyser et al. reported that females with SLE had low aerobic capacity when engaging in routine activities, leading to exhaustion even during light intensity activities of daily living.²¹ In a subsequent study, these investigators reported that VO₂ on-kinetics were prolonged even during the lowest energy expenditure levels associated with activities of daily living.¹⁴ Reduced cardiorespiratory fitness among patients with SLE has been reported by several additional studies,^22–24 with a recent study demonstrating poor cardiorespiratory fitness being directly related to performance fatigability in physically inactive females with SLE.⁵

VO₂ on-kinetics is a measure of the rate at which steady state VO₂ is achieved. When the cardiorespiratory system responds slowly to abrupt changes in energy requirements, anaerobic metabolic support is necessary to sustain the energy demand of the work being performed. By-products of anaerobic metabolism contribute to fatigue,⁹ and the transient increase in VO₂ to steady state is an indicator of fatigability. Faster VO₂ on-kinetics have been shown to occur with higher levels of cardiorespiratory fitness, while conversely, individuals who are untrained or physically inactive have VO₂ on-kinetics that are prolonged.^{10, 11, 25} Patients with cardiac or pulmonary dysfunction have also been observed to have slower VO₂ on-kinetics compared to control healthy individuals.^{12, 13} In females with SLE, VO₂ on-kinetics has been reported to be slower compared to a group of sedentary but otherwise healthy females, and implicated as a mediator of fatigue and fatigability that is prominent.¹⁴ The current findings were in agreement with these studies.

Aerobic exercise training has been previously shown to improve cardiorespiratory fitness among those with SLE,^26–29 with improvements being associated with reductions in patient reported fatigue severity²⁶ and increased physical activity.²⁷ These studies suggest that AET may be a sufficient stimulus to elicit a cardiorespiratory adaptation in females with SLE.

Following a 12-week aerobic exercise training program, we observed faster VO₂ on-kinetics among these females with SLE, consistent with reduced performance fatigability previously reported in these same subjects.¹⁷ VO₂ on-kinetics became faster by 37% on average, which was similar to changes reported by other studies of similar training duration.^{6, 30, 31} The magnitude of change in VO₂ on-kinetics was also related to participants’ initial status, whereby those with the slowest VO₂ on-kinetics had the largest improvement. This observation was consistent with studies in healthy individuals who did not have SLE,³² as well as other clinical populations.^{6, 30}

The adjustment of HR and [HHb] (representing exercise training-induced central circulatory and peripheral adaptations, respectively),²⁰ were examined in a subset of participants to acquire additional insight into the accompanying changes for VO₂ on-kinetics that are specific to those with SLE. The HR on-kinetics became faster with AET, while the time course of [HHb] only showed small changes. This suggests central O₂ delivery accompanied the changes in VO₂ on-kinetics response with exercise training in these females with SLE. This observation is in agreement with studies in healthy older and younger adults which have shown improved central O₂ delivery with exercise training.^{11, 19, 31}

For further insight into the components underlying VO₂ on-kinetics at the microvascular level, we expressed the magnitude and time course for [HHb] and VO₂ as a ratio to reflect the match between O₂ being provided and utilized by the active muscle.²⁰ Among healthy individuals, faster VO₂ on-kinetics is associated with smaller ratios of [HHb]/VO₂,³³ reflecting more efficient matching of O₂ distribution to O₂ utilization. For untrained men with slow VO₂ on-kinetics, large [HHb]/VO₂ ratios were observed together with an initial “overshoot” in the [HHb]/VO₂ ratio.³⁴ warm-up exercise however, this overshoot was abolished,³³ as a primed system provided better matching and negated the need for over-reliance of O₂ extraction to compensate for the increase in O₂ requirements.^{20, 33} We observed a similar overshoot occurring in these females with SLE, that remained even after exercise training, and the mean ratio of [HHb]/VO₂ did not change substantially after exercise training. This is in contrast to studies in young males¹⁹ and females³¹ that demonstrated complete abolishment of the overshoot and reduced ratio of [HHb]/VO₂ following exercise training. A similar observation was also reported after patients with type 2 diabetes engaged in moderate or high-intensity exercise training.³⁵ Interestingly, the overshoot persisted among older males¹⁹ and females³⁶ despite an improvement in the mean ratio of [HHb]/VO₂ following exercise training. This may suggest some limitations to microvascular blood flow remained in older but otherwise healthy indivduals.¹⁹ Within the literature, it has generally been observed that healthy females that were either trained^{31, 36–38} or underwent exercise training³⁶ showed more efficiency than their untrained self or counterparts, with faster VO₂ on-kinetics associated with lower ratios of [HHb]/VO₂ (Supplement Figure 3). The findings in this study are generally in agreement with these observations.

Cardiovascular disease^{39, 40} and cardiac dysfunction⁴¹ have been reported among patients with SLE. In general, increased physical activity and exercise has been reported to improve cardiovascular risk,⁴² endothelial function⁴³ and chronotropic incompetence,⁴⁴ and is often recommended as an adjunct to therapeutic regimens. Endurance exercise training has been shown to improve VO₂ on-kinetics in healthy participants as well as improve oxygen delivery.^45–47 At the microvascular level, dysfunction of the basement membrane affecting O₂ extraction in patients with SLE has been proposed, which may attenuate further improvement with exercise training. This is consistent with previous observation of muscle samples collected from patients with SLE.^{48, 49} Many other factors can influence microvascular oxygen delivery in females with SLE. Physical inactivity resulting in peripheral muscle deconditioning, glucocorticoid use, cardiac autonomic dysfunction, as well as many other disease related factors may adversely affect muscle oxygen diffusion in those with SLE leading to impaired kinetics.⁵⁰ While further investigation elucidating mechanisms of changes in microvascular oxygen delivery are needed in the SLE population, this study may contribute to expanded inquisition regarding potential microvascular and muscle oxygenation dysfunction.

Limitations

This is a pilot study that recruited only a small sample size with homogeneous demographics. The pre-experimental design did not incorporate a control group of either non-exercising females with SLE or a healthy reference group. Thus, neither the study observations nor their magnitudes may be generalized to the SLE population at large. The CWRT in this study utilized an active recovery period rather than passive rest or separate days for repeating the CWRT. This could result in some imprecision when comparing our results with studies using healthy individuals and precise VO₂ on-kinetic values reported by others. Possible un-intended effects of repeated bouts of exercise on VO₂ on-kinetics were also unknown, as were potential variations in physiological responses to AET related to phases of the participants’ menstrual cycle. The CPET and CWRT were performed in the morning for most participants, however, some of the participants had conducted or repeated these tests in the afternoon due to scheduling conflicts. The diurnal and day-to-day variability in this clinical population may have limited effect on the results of this study. The effect of the participants’ medication regimens on VO₂ on-kinetics, performance fatigability and cardiorespiratory fitness measurements was also unknown. The restricted range of the NIRS signal to penetrate tissue, especially in those with higher adipose tissue thickness, limits its application in clinical populations. Skin pigmentation is also known to affect the NIRS signal. It is unknown whether a different response would be expected in the exploratory outcomes, as these analyses excluded females with the highest BMI (range 31 to 38 kg/m²) and the degree of pigmentation was not accounted for. Use of HR on-kinetics as a proxy for central adaptations, as well as NIRS derived [HHb] data to represent local O₂ extraction are non-invasive measures causing minimal discomfort to the participants. However, an inherent limitation exists with the use of these devices to infer changes for outcomes that could not be directly measured in human studies.

Conclusion

This study evaluated changes in VO₂ on-kinetics in females with SLE before and after AET. Findings from this study demonstrated faster VO₂ on-kinetics, with the time course related to outcomes of cardiorespiratory fitness, in these females with SLE after 12 weeks of supervised AET. Faster HR on-kinetics accompanied the changes in VO₂ on-kinetics after AET, suggesting some central O₂ delivery adaptation occurring with exercise training in SLE.