Effects of arm cranking exercise on muscle oxygenation between active and inactive muscles in people with spinal cord injury

Abstract

Objective: We investigated the effects of the incremental arm-cranking exercise (ACE) on tissue oxygen saturation (StO₂) between active and inactive muscles, and the relationship between peak oxygen uptake (VO_2peak) and changes in the StO₂ in inactive muscles.

Design: Observational study.

Setting: Community-based supervised intervention.

Participants: The participants were individuals with motor and sensory complete spinal cord injury (complete SCI; n = 8) and motor complete but sensory incomplete SCI (incomplete SCI; n = 8), and able-bodied (AB) individuals (n = 8) matched for age, height, and body mass index.

Intervention: The ACE was performed at a rate increasing by 10 watts min⁻¹ until exhaustion.

Outcome Measures: VO_2peak, heart rate (HR), and StO₂.

Results: While VO_2peak was similar among the groups, peak HR was significantly higher in both SCI groups than in the AB (P < 0.05). In active muscles (biceps brachii), no differences in the StO₂ were observed among the groups (P > 0.05). In inactive muscles (vastus lateralis), the StO₂ in the AB and the incomplete SCI began to decrease at approximately 40% of the peak work rate; however, they remained unchanged in the complete SCI. The reductions in StO₂ in the AB were significantly greater than in the incomplete SCI.

Conclusions: These results suggest that sympathetic vasoconstriction occurred in the incomplete SCI and AB, although it did not occur in the complete SCI, probably due to a reduction in sympathetic nerve activity. Sympathetic vasoconstriction in inactive muscles may not contribute to an individual’s VO_2peak regardless of their group.

Introduction

Increased aerobic capacity as determined by peak oxygen uptake (VO_2peak) is strongly associated with a reduced risk of cardiovascular disease.^1,2 Based on Fick’s equation,³ VO₂ is the product of systemic flow (cardiac output) and oxygen extraction (a-vO₂ difference). Thus, systemic flow and O₂ extraction are important factors for increasing VO₂; one or both must increase to enhance VO₂. A study demonstrated that regular exercise induced an increase in cardiac output robustly elevate aerobic capacity.⁴ At high-intensity exercise, sympathetic nerve activity is enhanced,⁵ resulting in decreased blood flow to inactive muscles.⁶ In exercising skeletal muscle, however, local accumulation of metabolites blunts the sympathetically driven vasoconstriction in inactive muscles allowing for a marked increase in blood flow to active muscles, indicating an important role of sympathetic control.⁷ Indeed, one study in which participants underwent incremental leg cycling on electronically braked cycle ergometer (25 watts/min increase until participants were unbale to maintain a pedaling rate of 60 rpm) reported that greater reductions in tissue oxygenation in inactive arm muscle may cause a higher VO_2peak.⁸

Spinal-cord-injury (SCI) patients with paralysis of the lower limbs are a unique model due to their lack or lower sympathetic nerve activity.⁹ Abnormal heart rate (HR) and blood pressure are frequently observed in individuals with SCI; these responses are varied and depend on the extent of their injury levels.^10,11 Thus, these physiological variables, especially HR responses, may not be accurate measures to consider when determining aerobic capacity in individuals with SCI. Alternatively, focusing on skeletal muscle oxygenation may provide a better tool for evaluating aerobic capacity in these populations. Decreased muscle oxygenation, as an indirect indicator of sympathetic vasoconstriction in inactive muscles,^8,12,13 may not occur in persons with complete motor and/or sensory SCI. Conversely, one paper demonstrated that pressor responses were greater in tetraplegic participants compared to able-bodied control participants following noradrenaline infusion.¹⁴ These results mean potential exaggerated adrenergic receptor responses in tetraplegia, such that sympathetic vasoconstriction-induced reductions in muscle oxygen saturation in SCI may occur.

However, no studies have examined the effect of upper body exercise on muscle oxygenation in the paralyzed limbs of SCI patients (i.e. inactive muscles of their lower limbs during arm-crank exercise). Although it is supposed that muscle oxygenation may not be affected in individuals with both motor and sensory complete SCI, leg arterial blood flow assessed by venous occlusion plethysmography markedly reduced by skin stimuli, suggesting that occurrence of a sympathetic vasoconstriction, in individuals with SCI.¹⁵ Thus, it is possible that muscle oxygenation may be affected in individuals motor complete but sensor incomplete SCI. However, the potential differences in oxygenation between people with sensory complete and incomplete SCI are unknown. Periodic exercise training-induced an increase in VO_2peak is clinically important to reduce the risk of cardiovascular disease in SCI patients. For example, 10 weeks of arm-crank exercise training improved VO_2peak were associated with decreased body mass, triglycerides, and plasminogen activator inhibitor 1.¹⁶ Similarly, 12 weeks of arm-crank exercise training significantly improved VO_2peak and decreased plasma levels of leptin, tumor necrosis factor-alpha, interleukin-6, and waist circumference.¹⁷ Moreover, a significant relationship was found between improved VO_2peak and insulin sensitivity.¹⁸ Given these, further insight into the relationship between aerobic capacity and inactive muscle oxygenation responses is required, even for individuals with SCI with less metabolically active tissues.

We hypothesized that muscle oxygenation in lower limbs (inactive muscles during arm-crank exercise) in complete SCI may remain unchanged, while that in incomplete SCI and healthy controls may decrease during high intensity upper body exercise. Moreover, reduced muscle oxygenation in the incomplete SCI and healthy controls may be related to an individual’s VO_2peak. To test these hypotheses, we assessed near infrared spectroscopy (NIRS) signals during incremental arm-cranking exercise (ACE) until exhaustion in the complete and incomplete SCI as well as the able-bodied (AB) participants. Additionally, we evaluated muscle deoxygenation (HHb) derived from NIRS in active muscles, as HHb changes have been considered a surrogate for muscle O₂ extraction (i.e. a-v O₂ differences).^19,20 We aimed to estimate muscle O₂ extraction since it is one of the components of VO₂ that may be a more reliable variable than HR in SCI.

Methods

All procedures in the present study were approved by the ethical committee and were conducted in accordance with the guidelines of the Declaration of Helsinki.

Participants

Men with both motor and sensory complete SCI (complete SCI; n = 8) and motor complete but sensory incomplete SCI (incomplete SCI; n = 8), and eight age-matched AB participants were recruited for the study. Based on a previous study, SCI participants were recruited with spinal cord lesion below T7 as blood flow redistribution was affected by spinal cord lesion above T6, probably, due to splanchnic sympathetic control.²¹ After a detailed description and explanation of the study procedures, including the possible risks and benefits, each participant gave written informed consent and participated in the study voluntarily. The SCI participants were engaged in sedentary work to which they commuted using their own vehicle, and their daily wheelchair activity levels were similar. The AB participants were relatively sedentary and had not performed any regular exercise for 6 months before the study. All participants took no medications and did not smoke. The participants were asked to abstain from caffeinated beverages for 12 h, and from strenuous physical activity and alcohol for at least 24 h before the study. The participants were familiarized with the measurement techniques used in the ACE. All trials were conducted in a custom-made environmental chamber (Maeda Electrical Works Co, Ltd., Sapporo, Japan), maintained at a temperature of 23 °C with a relative humidity of 50%. The physical characteristics of the three groups are shown in Table 1.

Table 1. Physical characteristics among the three groups.

Participant No.	Class	Injury level	Injury duration (y)	Age (y)	Height (cm)	BM (kg)	ATT arm (mm)	ATT thigh (mm)
Complete SCI
1	A	T8	10	34	178	62.3	3.6	11.2
2	A	T10	14	37	176	63.8	5.1	10.5
3	A	T12	11	30	176	69.1	4.2	10.3
4	A	T8-9	28	55	165	71.5	5.1	8.5
5	A	T12	7	26	174	58.4	3.8	7.4
6	A	T9-10	7	26	171	62.6	5.3	7.8
7	A	T12	6	28	170	55.5	3.4	9.2
8	A	L1	8	35	176	65.7	3.8	8.3
Mean ± SD	NA	NA	11.4 ± 7.2	34 ± 9	173 ± 4	63.6 ± 5.2	4.3 ± 0.8	9.2 ± 1.4*
Incomplete SCI
9	B	L1	19	48	165	68.2	4.3	9.6
10	B	T9-10	9	35	174	52.5	5.2	11.5
11	B	T10	11	31	173	66.8	4.7	8.2
12	B	L1	15	41	166	51.3	3.1	10.2
13	B	T12	4	24	160	53.5	2.9	10.8
14	B	T12	6	32	182	65.2	4.5	9.7
15	B	T12	16	39	170	72.4	5.2	8.2
16	B	T10-12	6	28	172	69.8	3.4	8.8
Mean ± SD	NA	NA	10.8 ± 5.4	35 ± 8	170 ± 7	62.5 ± 8.6	4.2 ± 0.9	9.6 ± 1.2*
Able-bodied	NA	NA	NA	33 ± 4	172 ± 5	61.5 ± 6.5	4.5 ± 0.9	4.7 ± 1.2

Values are shown as mean ± standard deviation (SD). SCI, spinal cord injury; Class was determined by the standards of the American Spinal Injury Association. Complete SCI classified as A, SCI with no motor or sensory function preserved; Incomplete SCI classified as B, SCI with no motor but sensory function preserved; L, lumbar; T, thoracic. NA, not applicable; BM, body mass; ATT, adipose tissue thickness. * indicates a significant difference compared to the able-bodied group.

Incremental arm-cranking exercise

The incremental ACE was performed using an arm-cranking ergometer (Rehab Trainer 881E, Monark, Stockholm, Sweden). After 30 min of rest in the supine position, the SCI participants moved to their wheelchair and the AB participants moved to a wheelchair that was suitable for their body size. The wheelchair was firmly fixed, and the participant’s legs and feet were held in place with a strap. The pedal axis was aligned with the participant’s shoulder, with the elbow in a slight flexion position. Gas mask, electrodes for NIRS, electrocardiogram, and electromyogram were attached. It took about 20 min, and this period was enough time to avoid initial hypotension. The participants took a 10-min rest on a wheelchair and performed the ACE until exhaustion. The ACE, preceded by 4-min unloaded cranking, was performed at a rate of 10 watts/min at 50 rpm.¹⁶ The criteria for exhaustion were as follows: (1) failure to maintain 50 revolutions/min despite strong verbal encouragement, (2) a score of 19 on perceived exertion, and (3) respiratory gas exchange ratio at least >1.1. The test was terminated when the priority order from (1) to (2) and (3) was decided.^22,23

Anthropometric characteristics

Height, body weight, and adipose tissue thickness were measured while they were wearing only underwear and after they voided their bladder. Height was measured using a non-elastic tape measure in a 30° supine position, as this approximated the measurement of the standing AB participants.²⁴ Body weight of the SCI patients was measured using a custom-made scale. For the AB participants, body weight was measured using a commercial body weight scale (TBF-410, TANITA, Tokyo, Japan). Adipose tissue thickness was measured at both the biceps brachii and vastus lateralis in all participants using B-mode ultrasound (NEMIO SSA-550A, Toshiba, Tokyo, Japan).

Gas exchange variables

Pulmonary ventilation and gas-exchange variables were measured at rest and during exercise using an online computerized breath-by-breath method (V-max, Nihon Kohden, Tokyo, Japan). HR was monitored continuously via an electrocardiogram attached to a metabolic measuring system.

Near infrared spectroscopy signals

Local tissue oxygenation profiles of the biceps brachii muscle (active muscle) and the vastus lateralis muscle (inactive muscle) were measured using NIRS (BOM-L1TRW, Omegawave, Tokyo, Japan), as previously described.^12,25 This instrument uses three laser diodes (780, 810, and 830 nm), and calculates the relative levels of oxygenated hemoglobin (HbO₂) and HHb in the tissue according to the modified Beer–Lambert law. The total Hb was calculated as the sum of the HbO₂ and HHb, and the oxygen saturation of the skeletal muscle (StO₂) was expressed as HbO₂/total Hb as a percentage. NIRS optodes were placed over the belly of the biceps brachii muscle on the upper left arm of each partcipant,¹³ and on the lower third of the vastus lateralis muscle (10–12 cm above the knee joint).²⁶ The probe holder contained one light source probe, and two detectors were placed 2.5 cm (detector 1) and 4 cm (detector 2) away from the source at the vastus lateralis. Hb concentrations received by detector 1 were subtracted from those received by detector 2 at the vastus lateralis and 1.5 cm (detector 1) and 3 cm (detector 2) at the biceps brachii. This procedure minimized the influence of skin blood flow.²⁷ Since NIRS signals can reach half the depth of the distance between the probe and detector,²⁸ it can traverse at a depth of between 12.5 and 20 mm (thigh) or between 7.5 and 15 mm (arm). The NIRS signals were measured at 1-s intervals throughout the experiment.

Electromyogram activity

In order to confirm inactive muscle activities, the electromyogram was measured in AB participants during the ACE. The electromyogram was recorded for the vastus lateralis and rectus femoris using surface electrodes with a diameter of 5mm and an interelectrode distance of 20mm (EMG DDL -2A, Harada Hyper Precision, Sapporo, Japan), as previously described.^25,29 The signals were stored with a sampling frequency of 1 kHz. We expressed the electromyogram activity relative to the electromyogram activity during a maximal voluntary contraction. In SCI participants, no spasm was observed during the incremental ACE.

Data analysis

The breath-by-breath gas data were averaged every 15-s, and the highest 30-s data was determined as the VO_2peak. Similarly, the NIRS data were averaged over the last 30 s at each work rate to minimize muscle contraction-induced artifact of NIRS signal waves, based on our recent study.¹² To compare HHb in the active muscles of participants, HHb was expressed as a ratio of the resting baseline value. Because our NIRS device represented the HHb signal with arbitrary unit, the resting baseline values were defined as 100% and the differences were shown as relative changes.¹² A study reported that adipose tissue thickness has a large impact on NIRS signals, in that greater adipose tissue thickness means lower light absorption, resulting in lower signal changes.³⁰ To correct for this and enable comparisons between populations, a physiological calibration has been proposed.³¹ Since adipose tissue thickness of the thigh muscle was different between AB and SCI participants, we used this technique. A thigh cuff was inflated to 300 mmHg for 10 min after the incremental ACE test using a rapid cuff inflator (E20, Hokanson, Bellevue, WA, USA), as previously described.³² Changes in StO₂ were quantified as the total labile signal, defined in each experiment as the difference between the baseline (100%) and complete deoxygenation induced by muscle ischemia (0%). Thus, we represented StO₂ changes in inactive thigh muscles as changes from baseline (Fig. 1). The arterial cuff occlusion technique was not performed for active muscles because similar upper arm (active muscle) adipose tissue thickness existed among the groups. Moreover, we could not perform it for the exhausted muscles just after exercise due to ethical problems and claims of severe nausea and pain from the subset of participants during the preliminary test in our laboratory. All of the NIRS signals were plotted every 10% work rate of the peak VO₂ to compare the groups because an individual’s exhaustion time was different.

Figure 1.

Arterial occlusion test for inactive thigh muscles just after exercise of a representative subject. StO₂, tissue oxygen saturation; BL, baseline. The participant was able-bodied people.

Statistical analysis

A one-way analysis of variance (ANOVA) was used to compare the physical characteristics and the resting and peak cardiorespiratory variables among the groups. A two-way repeated measures ANOVA (work rate × groups) was used to compare the NIRS signals. Pearson’s correlation coefficient was used to examine the relationship between the VO_2peak and the StO₂ changes in the inactive muscles. Data are presented as mean ± standard deviation (SD). P < 0.05 was considered statistically significant.

Results

Physical characteristics

There were no significant differences in age, height, body weight, or adipose tissue thickness in the biceps brachii among the three groups, and no differences in injury duration between the complete and incomplete SCI groups. However, the adipose tissue thickness of the vastus lateralis in both SCI groups was significantly greater than in the AB participants (P < 0.05, Table 1).

Electromyogram activity

For AB participants, during the incremental ACE, muscle activity in vastus lateralis and rectus femoris muscles slightly changed from rest to at exhaustion (1.0 ± 0.2 to 3.6 ± 0.9% and 0.9 ± 0.2 to 3.1 ± 0.5% of maximal voluntary contraction, respectively). These values are quite similar as in a previous study that using the similar technique.²¹

Cardiorespiratory variables

At the resting baseline, there were no significant differences in any of the cardiorespiratory variables among the three groups. Similarly, the peak values of the gas exchange variables and the peak work rate were not statistically different among the three groups. Meanwhile, HR_peak in both SCI groups was significantly higher than in the AB participants (P < 0.05, Table 2).

Table 2. Cardiorespiratory variables at baseline and peak among the three groups.

	AB	Complete SCI	Incomplete SCI
Rest
VO2, ml min−1	316 ± 55	304 ± 64	301 ± 32
RER	0.88 ± 0.05	0.87 ± 0.02	0.89 ± 0.03
HR, bpm	70 ± 6	71 ± 8	69 ± 7
Peak
VO2, ml min−1	2027 ± 257	2023 ± 283	2004 ± 306
RER	1.14 ± 0.03	1.13 ± 0.02	1.11 ± 0.06
HR, bpm	149 ± 7	160 ± 8*	159 ± 8*
Work rate, watts	101 ± 18	105 ± 22	101 ± 11

Values are shown as mean ± SD. AB, able-bodied; VO₂, pulmonary oxygen uptake; RER, respiratory gas exchange ratio; HR, heart rate; bpm, beats per minute. * indicates a significant difference from the AB group.

NIRS signals

Figure 2A shows the active muscle HHb changes in all groups. The HHb kinetics showed a sigmoid curve upward to the higher work rate in all groups. In contrast, StO₂ in the active muscles was stable until the 20% work rate, and thereafter decreased linearly until the 100% work rate (Fig. 2B). In both NIRS signals, a two-way repeated ANOVA revealed no main effects of the groups and interaction, while a significant main effect of the work rate was observed (P < 0.05).

Figure 2.

Changes in muscle deoxygenation (HHb; panel A) and muscle oxygen saturation (StO_2; panel B) in the active muscles during the incremental arm-cranking exercise test among the three groups. Values are shown as mean ± standard deviation (SD). SCI, spinal cord injury; AB, able-bodied.

The inactive muscle StO₂ responses were different between the groups (Fig. 3). StO₂ in AB began to decrease at approximately 30% work rate, and the rate of the StO₂ decrease became steeper from approximately 70% work rate until exhaustion. In the incomplete SCI participants, StO₂ began to decrease from approximately 50% work rate and continued to decrease until exhaustion; however, the magnitude of the rate of the decrease was lower than that of the AB participants. In contrast, the StO₂ in the complete SCI group remained unchanged throughout the ACE. Significant main effects of the groups, work rate, and interaction were observed (all P < 0.05). A Bonferroni post-hoc test revealed that the StO₂ in the AB participants was significantly lower than in the complete SCI participants above 50% work rate and lower than in the incomplete SCI participants above 90% work rate. Moreover, there were significant differences in StO₂ between the incomplete and complete SCI participants above 80% work rate.

Figure 3.

Changes in the StO₂ in inactive muscles during the incremental arm-cranking exercise test among the three groups. Values are shown as mean ± SD. Symbols indicate the same as shown in Figure 2. * P < 0.05 between the AB and complete SCI participants, † P < 0.05 between the incomplete and complete SCI participants, and ‡ P < 0.05 between the AB and incomplete SCI participants.

Figure 4 shows the relationship between an individual’s VO_2peak per body weight and the inactive muscle StO₂ changes from the baseline to minimum values when all participants were pooled. There was no significant relationship between these two parameters in the analyses of all participants, AB and incomplete SCI, or AB alone.

Figure 4.

Relationship between an individual’s VO_2peak and changes in StO₂ in the inactive muscles during the incremental arm-cranking exercise test. Symbols indicate the same as shown in Figure 2. The three regression lines indicate the results of different groups.

Discussion

The major findings of the present study were: (1) HHb and StO₂ in active muscles during ACE were similar across the groups; (2) StO₂ in inactive muscles in the complete SCI participants remained unchanged, while those in the AB and incomplete SCI groups decreased above 30% and 50% work rate, respectively; and (3) no relationship was observed between the VO_2peak and the decrease in StO₂ in the inactive muscles.

Cardiorespiratory variables and HHb in active muscles

Previous studies reported that SCI patients compensated for lower stroke volume by increasing HR during ACE.^33,34 As no differences in the VO_2peak and HHb in active muscles were observed among the groups, our results are in line with these previous studies and are reasonable based on Fick’s equation (VO₂ = stroke volume × HR × a-v O₂ differences; HHb representing O₂ extraction).

A lack of differences in HHb and in StO₂ changes across the groups may indicate that muscle O₂ extraction was similar. It is generally accepted that oxygen extraction improves to some degree with exercise training.^35–37 Previous studies have also demonstrated that the rise in HHb is greater during incremental leg cycling^38,39 and HHb kinetics at the onset of exercise accelerate during constant-load exercises⁴⁰ to some degree with continuous exercise training. Furthermore, reductions in the StO₂ were greater compared with those pre-training.³⁹ Notably, the participants in these studies performed high-intensity exercise training, i.e. 70–100% work rate.^35,38–40 People with SCI normally use a wheelchair in daily life, suggesting that upper body physical activity in SCI patients may be more frequent compared to AB individuals. However, our results suggest that the daily wheelchair use of SCI patients may be insufficient to improve muscle O₂ extraction, resulting in no differences in the NIRS metrics between the SCI and AB participants in this study. Indeed, previous studies reported that electrical stimulation in the form of twitches may lead to improvement in aerobic capacity of the muscle following training in persons with complete SCI⁴¹ and that evoked resistance training may result in muscle hypertrophy, potentially leading to improvement of muscle O₂ extraction.⁴² Thus, our results may be supported with these findings.

Sto₂ in the inactive muscle

We used the StO₂ calculated by dividing HbO₂ by the total Hb as an indicator of sympathetic vasoconstriction,¹² since HbO₂ is strongly affected by skin blood flow,⁴³ and total Hb represents blood volume.¹² Thus, StO₂ may be a reliable indicator of tissue O₂ saturation (or arterial inflow to the target muscles) with a minimum effect of the skin blood flow and blood volume. StO₂ decreased in accordance with an increasing work rate in the AB and incomplete SCI participants; however, in the complete SCI participants, it remained unchanged. Moreover, the magnitude of the decrease in StO₂ was different between the AB and incomplete SCI participants. Several possibilities account for this. Stjerberg et al. (1986) recorded muscle sympathetic nerve activity in SCI patients below the level of injury, and reported that spontaneous muscle sympathetic nerve activity was much lower in SCI patients than control subjects.⁹ Lower sympathetic nerve activity can diminish vasoconstriction in paralyzed lower limbs. Previous studies have demonstrated that reductions in muscle oxygenation may represent sympathetic vasoconstriction in both active^44,45 and inactive muscles.^12,13 Our results indicate that sympathetic vasoconstriction in inactive muscles did not occur in the complete SCI participants, but occurred at approximately 40% work rate in the incomplete SCI and AB participants (Fig. 3), which is consistent with a previous study.⁴⁶ However, the different magnitudes of StO₂ (i.e. reduction rate) between the incomplete SCI and AB participants should be considered. Below the level of SCI lesions, reduced peripheral capillarization⁴⁷ and decreased conduit artery diameters have been observed.^48,49 As the NIRS signals traverse directly under the probe and detectors and measure whole vascular beds, our results may be related to the lower blood volume in paraplegics’ lower limbs.

Relationship between VO_2peak and inactive StO₂

Contrary to our hypothesis, we did not find any significant relationship between VO_2peak and decreases in StO₂ in the inactive muscles. These results may indicate that sympathetic vasoconstriction in inactive muscles at high exercise intensity may not contribute to blood flow redistribution to active muscles, which is inconsistent with our previous study.⁸ One reason for such discrepancy may be the smaller muscle mass recruited during arm-cranking, leading to different exhaustion levels among individuals.

The foundational premise of VO_2max, namely that there is a speed of locomotion or rate of work above which VO_2max fails to increase further, was proposed in 1923 by Hill and Lupton.⁵⁰ To date, secondary criteria of VO_2max have been reported, that is, a blood lactate concentration >8 mM, or a respiratory exchange ratio (RER) >1.10.^51,52 In the present study, although the peak RER during the ACE appeared to reach the criteria of VO_2max on average, the validity of HR_peak should be considered. According to Tanaka’s formula,⁵³ the HR_peak values in this study were 87% and 81% of the age-predicted values in the SCI and AB participants, respectively. This formula was derived using large leg muscle mass during exercise.⁵³ A study demonstrated that this formula overestimates age-predicted HR_peak by approximately 15 bpm compared with the true HR_peak during upper body exercise.⁵⁴ Taking this into consideration, the HR_peak in this study may reach 90–95% of the age-predicted maximal HR during ACE. It should be note that abnormal HR responses have been observed in individuals with SCI, indicating an inaccuracy for evaluation of HR_peak in these populations.^10,11

However, active muscle HHb showed a sigmoidal curve, resulting in a plateau phase at the maximal work rate that was observed in a study during leg cycling at exhaustion.⁵⁵ Therefore, it is possible that no relationship between the VO_2peak and StO₂ changes existed with the fatiguing workload imposed on the arms, without Q (HR × stroke volume) being maximal.

Study limitations

Several limitations should be considered. First, we could not measure sympathetic nerve activity directly using microneurography due to device and technical limitations. Also, this study included SCI with only low thoracic and lumbar motor complete SCI, not cervical or upper thoracic SCI. Recruitment of these participants locally was difficult, and IRB approval could not be obtained for these populations. Further, we could not completely rule out the effects of skin⁴³ and adipose tissue thickness^30,56 on NIRS signals. However, HHb as an index of muscle O₂ extraction was not strongly affected by skin blood flow. We also performed physiological calibrations, and used two detectors to minimize the effects of skin and adipose tissue thickness on the NIRS signals.³¹ A recent study debated the validity of NIRS signals in people with SCI, due to several methodological and physiological limitations (e.g. muscle atrophy, intramuscular fat, and movement artifact).⁵⁷ We must acknowledge the potential limitations of the device, and therefore, future studies are required.

Conclusion

During ACE, the NIRS metrics in active muscles were similar among the three groups. The inactive muscle StO₂ in complete SCI participants remained unchanged, while in AB and incomplete SCI participants it began to decrease above 40% work rate, possibly due to the effect of sympathetic vasoconstriction. There were no significant relationships between VO_2peak and StO₂ reductions in the inactive muscles when the data were pooled, indicating that sympathetic vasoconstriction in inactive muscles may not contribute to the peak aerobic capacity in the populations of this study.

Disclaimer statements

Contributors None.

Funding The study was supported by the Japan Society for the Promotion of Science (grant number 16700497).

Conflict of interest Authors have no conflict of interests to declare.