Introduction

A Spinal Cord Injury (SCI) causes physiological and functional disorders whose severity is more or less dependant on the level of injury. These disorders have an impact on the everyday life of paraplegics and often lead to a physiological deconditionning. Metabolic changes are also associated with the reduced activity following SCI. The limited activity often negatively affects the health of these people and leads to a debilitative cycle. Muscle atrophy could be observed, due to a decrease in the oxidative and glycolytic enzyme activity in arm muscles.1 A risk of atherosclerosis,2 a reduction of aerobic capacity3 and other cardiovascular consequences4 are linked with the lower basal metabolic rate of these patients. Furthermore, high density-lipoprotein (HDL) cholesterol levels are reduced in SCI patients.1 In fact, the limited life-style may lead to a physiological deconditioning, increasing the risk of coronary disorders, obesity and diabetes.5

In this regard, it appears that regular physical activity may be a decisive factor for the well-being of paraplegics. It has been observed that a rehabilitation program improves the cardioventilatory functioning.2,3,6,7,8

One of the major problems encountered during a rehabilitation program is the ability to measure the level of fatigue after a given effort, without the use of special physiological equipment. Since only a single intense load of training is required to provoke beneficial adaptations for the individual,5 it is important to have a measure of the strain a given exercise imposes on the individual to be sure the stimulus is sufficient to induce an adaptative response. Authors have tested perceived exertion (PE) and different rating scales in healthy people9,10,11,12 and people affected by different pathologies during a training or rehabilitation program.12,13,14 These rating scales permit subjects to control accurately the exercise intensity without using technical instrumentation (ie heart rate monitoring), that could be interesting for patients affected by SCI who do not have use of medical instrumentation. Previous studies have demonstrated that PE was linked with peripheral factors such as muscle pain, and cardioventilatory factors such as heart rate or oxygen uptake.15 However, the exact influence of these factors is still unknown in paraplegics. An investigation of deconditioned subjects such as paraplegics may offer decisive data on the processes linked with PE.

Therefore, the purpose of this study was to compare the effects of a rehabilitation program on the PE and the cardioventilatory responses during exercise in healthy people and paraplegics.

Materials and methods

Subjects

A group of seven healthy males (age 26.6 SD 6.2 years) and a group of seven old paraplegic males (age 35.2 SD 15.9 years, time since occurence of SCI 12.3 SD 10 years) volunteered to participate to this experiment after having been informed about the procedure, and signed an informed consent. Results from a standard health questionnaire, clinical and electrocardiogram examinations were normal for all subjects. According to the American Spinal Injury Association (ASIA) classification rules paraplegics were classified as ASIA A. The subject characteristics are presented in Table 1. All subjects were physically active without specific upper limb training. Our local ethical committee has approved this study. The experiments comply with the current laws of France.

Table 1 Characteristics of subjects of the two gruops
Full size table

Methods

All subjects performed a maximal graded exercise test (GXT)16,17,18,19 on an ergometer (VP100H–HEF tecmachine, Andrezieux Boutheon, France) using the personal wheelchair for the paraplegics and a classic wheelchair for the healthy subjects (seat dimensions 15 inches by 18 inches, seat height 30 inches) with the position always held constant.20 However, by means of a foam cushion, each subject could adjust their arm position. The ergometer had a single motorized roller with enslaved braking, and an electromagnetic system.21 This system was controlled by specific software which allows the subject to work at a pre-determined speed. During the exercise, subjects could see all of the information concerning PO and speed.

After 6 min of quiet sitting in the wheelchair ergometer, the subject performed a progressive exercise test at a constant speed and 30 rpm16,17 with no load. The load was increased from 0 W by 10 W every 2 min until exhaustion. The protocol of this test has been previously established by Predine et al16 after several trials at various loads (5, 10, 20, 30 watts), duration (1, 2, 3 min) and velocities (30, 60, 90 rpm). The same authors showed that this test is reproducible, allows a warm-up, and was in every case long enough to ensure the measure of the maximal tolerated power (MTP), peak oxygen uptake and maximum HR.

The highest load which could be maintained with a constant speed of the ergometer for 1 min was taken as the MTP. The end point of the test was determined when the subject could not maintain the expected cadence or power output.

Expired gases were sampled and analyzed every 30 s with an Oxycon Champion analyzer (Mijnhardt®), calibrated before each test. The respiratory parameters monitored included oxygen uptake (VO2, ml.min−1. kg−1. Heart rate (HR, bpm) was monitored continuously during the test by telemetry (PE 4000 sport tester). The perceived exertion and perceived pain were measured by the BORG CR 1012 immediately at the end of the GXT.

Training period

The training program consisted of three sessions per week for 6 weeks. Each session consisted of a 45 min SWEET for both groups proposed by Gimenez et al.19,20 This particular type of exercise, used work rates established from the ventilatory threshold (VT) and the MTP of the progressive test, consisting of successive work bouts of 5-min duration. During each work bout, a 4 min period of moderate work, termed `base' level, was followed by a 1 min period of heavy work, termed `peak' level. Initially the base is set at a work rate corresponding to the VT and the peak at the MTP. The maximum intensity of endurance (MIE45) was defined by both maximal HR at the end of the last peak and the inability for the subject to finish a training session. The peak and the base loads were alternately readjusted (+10 W) each time the heart rate at the end of the test was decreased by 10 bpm compared to the last increase. So, each training period was designed to be performed at the MIE45.

The maximal graded exercise tests and the SWEET were performed before (GXT 1 and SWEET 1) and after a 6-week training period (GXT 2 and SWEET 2). All tests, including the training session were performed under medical supervision. Attendance at the training sessions was excellent, every subject followed the experimental protocol and has fully realized the program training session. To assess the effects of training, the last training session (SWEET 2) was performed at the same work load as the first (SWEET 1). At the end of each SWEET, subjects pointed on the CR 10 scale to a number that corresponded with his or her perception of effort and pain.

Statistical analysis

Data were analyzed with a two-way analysis of variance (groups×measurements) and a Scheffé's post hoc test for individual differences. Values are presented as means (M) and standard deviation (SD).

Results

Maximal graded tests (Table 2)

Table 2 Mean values and standard deviation (SD) of perceived exertion, heart rate, maximal tolerated power and maximal oxygen uptake measured during the maximal graded exercises before (GXT 1) and after (GXT 2) a rehabilitation training program performed by paraplegics and healthy subjects
Full size table

Analysis of variance of PE revealed no significant main effects in either group between GXT 1 and GXT 2. A significant main effect for groups for Perceived pain (PP) was observed (F1, 11=5.037, P<0.05). Scheffe's post hoc test revealed a PP lower for paraplegics for the GXT 1 (M=3.25 SD 4.2) compared to the healthy subjects (M=5 SD 2) and for the GXT 2 (M=2.58 SD 2.2 and M=6 SD 1.6 respectively). However, no significant effect for measurement was observed.

Analysis of variance of maximal HR revealed no significant main effect in either group between GXT 1 and GXT 2.

A significant main effect for measurements of MTP (F1, 11=66.296, P<0.0001) was found. Scheffé's post hoc test revealed a significant increase for both groups between GXT 1 (M=73.7 W SD 12.6 for paraplegics and M=61.6 W SD 16 for healthy) and GXT 2 (M=88.1 W SD 18.2 and M=89.3 W SD 15 respectively). Furthermore, a significant main effect for group and measurement (F1, 11=6.589, P<0.05) was found. MTP of the healthy group was significantly higher than paraplegics before and after training and the increase of healthy group was significantly higher than the paraplegic group. Peak oxygen uptake significantly increased for both groups (F1, 11=7.951, P<0.05) from GXT 1 (M=27.73 ml.min−1.kg−1 SD 6.1 for paraplegics and M=34.66 ml.min−1.kg−1±4.9 for healthy) to GXT 2 (M=32 ml.min−1.kg−1 SD 6 for paraplegics and 37.54 ml.min−1.kg−1 SD 4.3 for healthy). However, a significant group effect (F1,11=5.171, P<0.05) was found. The maximal oxygen uptake was significantly higher in the healthy group for both tests.

SWEET Tests: (Table 3)

Table 3 Mean values and standard deviation (SD) of base heart rate, peak heart rate, perceived exertion and perceived pain measured at the beginning (SWEET 1) and at the end (SWEET 2) of a rehabilitation training program performed by paraplegics and health subjects
Full size table

A significant main effect for measurement of PE was observed (F1, 11=16.348, P<0.01). PE was significantly lower for SWEET 2 for both groups (M=2 SD 1.2 for paraplegics and M=2.5 SD 1.1 for healthy) compared to SWEET 1 (M=4.86 SD 2.5 and M=5.14 SD 1.3 respectively). However, no main effect of group was observed. A significant main effect for measurement of PP was observed (F1, 11=5.865, P<0.05). PP was significantly lower for SWEET 2 for both groups (M=1.04 SD 1.9 for paraplegics and M=0.5 SD 0.6 for healthy) than SWEET 1 (M=2.07 SD 2.9 and M=3.69 SD 2.1 respectively). Nevertheless, no main effect of group was observed.

A significant main effect of measurement was observed in the `base' HR (F1,11=36.059, P<0.0001). Post hoc test revealed a significant decrease between SWEET 1 for both groups (M=149.18 bpm SD 20.6 for paraplegics and M=145.71 bpm SD 21.8 for healthy) and SWEET 2 (M=135.06 bpm SD 20.2 and M=117.57 bpm SD 19.4 respectively). A significant main effect of measurement was found in `peak' HR (F1, 11=26.995, P<0.001). The peak HR was significantly higher in SWEET 1 for both groups (M=157.96 bpm SD 24.6 for paraplegics and M=164.71 bpm SD 18.8 for healthy) compared to SWEET 2 (M=146.71 bpm SD 25.9 and M=129.57 bpm SD 20.7 respectively). However, a significant interaction (F1,11=7.165, P<0.05) between groups and measurements was observed. The decrease of peak HR was higher for healthy than paraplegics.

Discussion

Firstly, the results must be interpreted with some precautions because of the relatively small size of both groups. Further investigations carried out on a large experimental population must confirm our results. Previous studies have reported that HR and perceived exertion were variables significantly correlated.9,10,11,15 The fact that no difference of PE was found between the two groups during the two GXT suggests that paraplegics of the present study felt the same perceived exertion as healthy people. Furthermore the results of this study show no significant difference between the two groups in peak HR between the two GXT and between the two SWEET tests. It appears as though both populations have the same heart rate responses during the rehabilitation program. It may be expected that the groups have different responses to the training, but these results may be possible because the paraplegic patients in this study were heterogeneous and the standard deviation of HR was greater, involving no significant difference.4 However, previous studies have reported in healthy subjects that perceived exertion can be used to control exercise intensity during continuous9,11,13 or intermittent training.22 Therefore, it may be suggested that exercise intensity could also be controlled by perceived exertion in paraplegic subjects without depending on any biofeedback monitoring system (ie heart rate monitoring).9,12

The fact that no significant difference was found in HR and PE between GXT 1 and GXT 2 indicated that maximal graded exercises were performed at the same relative intensity. The significant decrease of PE and HR values observed during SWEET 2 compared to SWEET 1 at the same power output demonstrates the positive effects of the rehabilitation training program on the physiological responses of subjects. Other authors have found similar findings.4,7

A difference in the perception of pain was observed between GXT 1 and GXT 2 between the two groups. The paraplegics seem to feel less pain compared to healthy subjects during exercise. This result may be explained by different past ways of life of each group.8 Indeed, paraplegics have a different approach of pain because of their spinal cord injury.14 Also, the difference of pain feeling could be explained by the fact that old injury paraplegics tested in the present study are trained for a long time to use their arms to move their wheelchair.

However, a difference in perception of pain in both groups was observed between the SWEET tests. This result further supports the positive effects of the rehabilitation program for both groups. It is likely that the rehabilitation program develops specific peripheral muscular adaptations against pain. The absolute work loads were the same between both tests, but the relative work loads were lower.

The results show significant changes in the maximal tolerated power and oxygen uptake in both groups between the two maximal graded tests showing the positive adaptations of the rehabilitation program. However, the training effects are more significant in the healthy people than in the paraplegics. The paraplegics and the healthy people have different adaptations of their thermoregulatory systems and of their arterial pressure during exertion.14

The thermal regulation disorders of paraplegics cause hypothermia at rest and hyperthermia at exercise that may create limitations in the metabolism and also diminishing sweat responses. The regulation of cellular and extracellular hydration is disturbed, altering the exercise capacity. These disorders may influence the arterial tension, which may result in a decreased exercise intensity supported by paraplegics. These diminished responses or lack of adaptations compared to healthy subjects arises directly from primary complications associated with the spinal injury. Moreover, the healthy subjects did not have experience of handrim propulsion in wheelchair and there improvement could be better than the paraplegics.23 Therefore, all these parameters suggest that the training effects are less significant for paraplegics than healthy subjects. Furthermore, the ventilatory responses of paraplegics may also be affected by decreased efficiency of the inspiratory muscles and their improvement may not be as great compared to the healthy subjects.3 However, it appears that these ventilatory disorders do not significantly affect perceived exertion of paraplegic patients, as it does with cardiac patients with chest pain.15 Finally, as oxygen uptake decrease with increasing age,24 it is possible that the significant changes in the maximal oxygen uptake observed between both groups could also be influenced by the age difference of the subjects. The time elapsed since occurence of SCI may also influence the rehabilitation capacities of patients. However, the fact that only old spinal cord-injured patients were tested in the present study does not allow to determine influence of this variable.

Conclusion

Our results suggest that the spinal cord injury does not influence perceived exertion of old SCI patients compared to healthy subjects. Their training capacities seem to be real, although the paraplegic patients were older than the healthy subjects. Thus, these findings support the use of perceived exertion as a useful measure of exercise intensity for the rehabilitation purposes of paraplegic patients. Further investigations are needed to determine the use of perceived exertion as a means of self-regulation of a rehabilitation program.