Comparison of muscular activity on ergometric bicycle and elliptical trainer in subjects with incomplete spinal cord injury

Introduction: Spinal cord injury generates muscle weakness, impairing orthostatism and gait. The elliptical trainer (ET) and the ergometric bicycle (EB) are rehabilitation options for this subject. Understanding the pattern of muscle activation generated by these methods is important to answer questions arising from clinical practice. Objective: To verify muscle activation with ET and EB with and without electromyographic biofeedback in subjects with incomplete spinal cord injury (ISCI). Methods: Cross-sectional crossover study, enrolled in Clinical Trials (NCT05118971). Subjects with spinal cord injury (incomplete spinal cord injury group - ISCIG) and without spinal cord injury (reference group - RG) were randomized into four groups: elliptical group (EG), elliptical + biofeedback group (EBG), bicycle group (BG) and bicycle + biofeedback group (BBG). Subjects were assessed for functionality by the Functional Independence Measure, injury classification by the ASIA Scale, muscle tone by the modified Ashworth scale, and muscle activity by electromyography. Results: There was greater activation of the tibialis anterior on cycling compared to other modalities in ISCIG. Biofeedback offered no difference in any of the groups. In RG the vastus medialis was the most activated muscle in all modalities, with more expressive activation in the ET. In this same group, the tibialis anterior was more activated on the EB. Conclusion: This study showed that both ET and EB are safe and effective in recruiting the muscles investigated, encouraging its use by rehabilitation professionals when the objective is to strength muscles involved in gait.


Introduction
Spinal cord injury (SCI) is caused by any harmful process in the spinal cord impairing its function, causing muscle weakness and orthostatism deficit.1,2 In incomplete spinal cord injuries (ISCI), the deficit in muscle activation compromises many functional actions, such as gait.Due to the involvement of many cortical, subcortical, and spinal neural centers in locomotor function, it is important that intervention strategies are designed to target neural elements at all levels of the neuroaxis.[3][4][5] Currently, there are encouraging results with physiotherapeutic to improve the prognosis of locomotion movement, 6,7 since the restoration of the ability to walk is an area of great interest in the rehabilitation of people with this type of injury.8 The trunk extensor muscles are also essential in this process, since it is largely responsible for maintaining standing posture.9 To assist in the locomotion of subjects affected by an ISCI, there are several resources available in clinical practice, such as parallel bars, crutches, and canes.
However, these tools require weight bearing and are not constant or easily quantifiable.Other devices widely used in clinical practice are the elliptical trainer (ET) and the ergometric bicycle (EB), still with few studies investigating its use in neurological patients.The first is a walking simulator, advantageous for reducing the impact on the joints, benefiting subjects with impaired balance, considering that the feet remain in continuous contact with the support surface.12 The EB, widely used in the motor rehabilitation, can also be a tool for muscle recruitment, facilitating the active movement of the knees during cycling, especially extension.12 Therefore, since there is: a) a high incidence of incomplete spinal cord injuries with significant gait  In all physical therapies modalities, after proper positioning, the subjects were asked to perform the movements at the maximum speed possible.
Before acquiring the signals, one minute warm-up was performed to learn the movement, and then the electromyographic signal of the selected muscles was recorded for 40 seconds.The same procedure was performed with the addition of biofeedback, and then they were instructed to follow the evolution of the graphs that represented muscle activation, aiming to increase their amplitudes with greater muscle recruitment.It should be mentioned that for those who were unable to support their own weight, it was necessary to place them The heel should make contact with the pedal of the stationary bike and the elliptical in the movement of the limb that would correspond with the single stance phase of gait, the knee should not be flexed, and the contralateral limb that would correspond with the swing phase of gait would be in a knee flexion of approximately 60º, which corresponds to the maximum flexion achieved in the entire cycle of a physiological gait 17 (Figure 1).
therapies modalities: EG, EBG, BG and BBG.On the first day, subjects performed the first modality drawn and later biofeedback was added.In cases where the first modality drawn was with the biofeedback, it was later removed.In the second meeting, the crosssover was made, evaluating the remaining modality.The interval between assessments was at least seven days and the participants were asked not to perform any exercise in the last two days preceding the meetings.
To determine the activity of the vastus medialis, gluteus medius, tibialis anterior and paraspinal muscles, surface EMG was used, as the EMG signal originates from the electrical activity of the motor endplate.
The signal is generated by the union of the activation potentials of the motor units that are concomitantly recruited, resulting in the interference tracing.

Results
For the ISCIG initially 13 individuals were selected according to the inclusion criteria, but TWO did not return for the second evaluation, totaling 11 subjects in the ISCI group.For the RG, 7 individuals were invited to participate, and all completed the evaluations.Table 1 shows the characteristics of all subjects who participated in this study.Table 2 shows the MAS results for the ISCIG.
To identify how much the SCI subjects present        Cycling provides activation and strengthens several muscles.The descending phase leads to activation of the gluteus maximus, gluteus medius, vastus intermedius, vastus lateralis, vastus medialis, rectus femoris, gastrocnemius and soleus, with hip and knee extension.
In the ascending phase, the ilipsoas, hamstrings and anterior tibialis muscles are activated with consequent hip and knee flexion and dorsiflexion.21 On the other hand, on the elliptical, the ankle does not satisfactorily perform plantar flexion and dorsiflexion, therefore, the activity of the muscles involved in this joint are Note: Correlations between muscle activation and FIM were measured using Spearman's correlation coefficient, where the R assumes a value between -1 (negative correlation) and 1 (positive correlation).The significance level adopted was *p < 0.05.There was a correlation between tibialis anterior (TA) muscle activation and FIM both on the bicycle and on the elliptical: the greater the stimulus, the greater the functional Our study had some limitations, such as heterogeneity and sample size.The studies available in the literature with this population have a similar sample size, as it is difficult to recruit these individuals due to locomotion, for example, which makes it even more difficult to obtain a homogeneous sample with similar injuries and deficits.
The imprecision related to the weight supported by the individuals, being positioned by the knee angle, but without numerical precision of the suspended weight was also considered a weakness.

Conclusion
Our study demonstrated greater activation of the tibialis anterior muscle in the cycling exercise both in individuals with ISCI and in the group of non-injured subjects compared to the elliptical.We also observed the close relationship of the tibialis anterior with functional activities through its correlation with the functionality scale, demonstrating the importance of this musculature in everyday life.Furthermore, for the RG, the vastus medialis muscle was the most activated in all modalities, showing its highest values in the elliptical modality.
Our results suggest that the use of the ergometric

10
Partial body weight support (PBWS), on the other hand, is one of the most used systems today, as it provides a decrease in the lower limbs load, facilitating movement, since body weight is reduced through suspension, facilitating movement for subjects who lack strength to support its own weight, 10,11 in addition to a significant decrease in the metabolic cost during walking.11 PBWS can be used in a treadmill, walker and even in the ground.In a study carried out by our research group, investigating which of these devices caused greater muscle activation in the main muscles involved in gait in subjects with ISCI, it was observed that treadmill training caused less fatigue and greater muscle activation when compared to gait training with a walker.10 Understanding the muscle activation pattern generated by these methods becomes important to answer questions arising from clinical practice.6 Muscle tone was assessed using the MAS, evaluating hip flexors and extensors, hip abductors and adductors, knee extensors, dorsiflexors and plantiflexors.The score is given as follows: 0: no increase in muscle tone; 1: increased tone at the beginning or end of the range of motion; 1+: increased tone in less than half the range of motion, manifested by abrupt tension and followed by minimal resistance; 2: increased tone in more than half of the range of motion; parts in extension or flexion and moved with difficulty; 4: Rigid parts in flexion and extension.13 Muscle activation was assessed through EMG.The subjects in the ISCIG and the RG were randomized by drawing lots in an envelope with brown paper and inserted in one of the four groups: elliptical (EG), elliptical + biofeedback (EBG), bicycle (BG) and bicycle + biofeedback (BBG).A person not involved in the research was asked to remove a paper, defining the order in which the subjects would perform the physical FISIOTERAPIA EM MOVIMENTO Physical Therapy in Movement Santos PP et al.Fisioter Mov.2023;36:e36131 4 In both modalities, when necessary, manual assistance was provided by the physiotherapists on each leg to assist in the movements for the ISCIG.The physiotherapist's hand was placed on the anterior surface of the leg below the patella to assist with knee extension during the phase corresponding to the single stance phase of gait.The other hand was placed on the ankle to help move the feet.Subjects with trunk control and ambulators performed both modalities without suspension and without assistance.
7,14 To acquire the EMG signal, an electromyograph was used (New Miotool Fisio, Miotec, Porto Alegre, RS), 2000Hz/ chanel, eight channels connected to the computer (Dell/ Windows 10) with the software Miograph (Miotec, Porto Alegre, RS), amplified acquisition at 2,000 Hz, noise < 2 LSB, common mode rejection ratio of 126 dB, input impedance of 10 Ohm //2pF and bandpass filter from 20 to 500 Hz.Skin impedance was reduced by shaving the area and with a gentle antiseptic abrasion with cotton soaked in 70% alcohol, based on International Society of Electrophysiology and Kinesiology guidelines.To capture the signal, disposable adhesive electrodes Meditrace 200 (Kendall, EUA) were used, measuring 10 mm in diameter and with bipolar configuration.The distance between the electrodes was 20 mm, as recommended by SENIAM (Surface ElectroMyoGraphy for the Non-Invasive Assessment of Muscles).The electrodes were positioned to allow the acquisition of the EMG signal in the following muscles: vastus medialis, gluteus medius, tibial anterior and paraspinal, with a reference electrode on the anterior tuberosity of the tibia.Acquisitions were performed on the hemibody with better muscle performance during strength assessments.After placing the electrodes, subjects who could not support their own weight without assistance were supported by the Hip Positioner (Pacer Gait Trainer, Rifton, Rifton, NY), which was attached by nylon straps to a galvanic iron grid fixed to the ceiling.Subjects were positioned on the PBWS with the help of two physiotherapists.The height of the seat belts and the bicycle seat was measured according to the patient's proper position based on the analysis of the gait pattern.

Figure 1 -
Figure 1 -Positioning on the elliptical with partial body weight support (A) and exercise bike (B).

Fisioter
Mov. 2023;36:e36131 5 in the PBWS to use the ET.As the EB had a seat, only manual support was given as needed.The studied population does not allow testing the maximum voluntary contraction, since they suffer from a severe decrease in strength and difficulty in selecting movements, making it impossible to normalize the electromyographic data.A formula constantly used in electrical signal analysis is the root mean square (RMS) which is performed by calculating the root mean squares, a feature offered by the application itself, which we use instead of traditional normalization.18The same EMG signal acquisition procedures were performed in the RG subjects, matched by mean age and gender.The purpose of the RG was to obtain a parameter of EMG values in healthy individuals.The same positioning measurements were performed on the devices as described.The side selected for the assessments was according to the subject's laterality.Regarding the normalization of the data, even though the procedure recommended for healthy individuals is the maximum voluntary contraction, in this study we used the calculation of the RMS, for proper comparison with the ISCIG.19,20Statistical analysisTo determine the sample size, we considered the difference in the mean between the EMG muscle activity of the rectus femoris muscle during 2 exercise modalities (steady walker and treadmill at 1 km/h, where the difference in the mean was 4.72 ± 1.13), with an effect power of 90% and a significance level of 0.05.Therefore, 11 subjects were calculated for the study.The calculation was performed in the GPower software, measured by the muscle activation variable based on the study carried out by Borin et al.10 We emphasize studies with similar methodology and with the spinal cord injured population are very scarce in the literature, so for that great reason the aforementioned methodology was used.The Shapiro-Wilk normality test was performed.The results of qualitative variables were presented through frequency and percentage.Quantitative results were presented as median and interquartile range (IQR), except age and FIM, presented as mean and standard deviation.Both for comparing the activation of the four muscle groups analyzed and for the four exercise modalities, Friedman´s test was used with Tukey´s post hoc.Correlations of muscle activation with age, duration of injury and FIM were measured using Spearman's correlation coefficient.The significance level adopted was 0.05.Analyzes were performed using the SPSS statistical software (IBM SPSS Statistics for Windows, 25.0, Armonk, NY).

a
smaller muscle activation when compared to the individuals of the RG, a normalization of the absolute values of the EMG of the ISCIG was carried out in relation to the RG.The medians of the RG were calculated in relation to each muscle/modality.Afterwards, it was generated a percentage of how much each muscle reached in relation to the normality for each exercise modality.Table3shows how the ISCIG values are lower when compared to the RG.

reduced. 22
Burnfield et al. 22 point out that the reduction and anticipation of the activity of the ankle muscles decrease the reflex activation of the leg stabilizers.The elliptical reduces the demand on distal muscle activity and promotes proximal muscle strengthening, 22,23 corroborating our findings.The greater activation of the tibialis anterior on the bicycle is an important finding due to its functional relevance in gait.Muñoz et al. 24 carried out a kinematic study of the feet and EMG of the tibialis anterior and bilateral gastrocnemius muscles during gait with different shoes in healthy individuals.When walking with bare feet, the authors observed higher activation of the muscles and the greater EMG activity was in the tibialis anterior, whose function is to perform dorsiflexion, also acting to cushion the impact with the ground during walking, emphasizing its importance for this functional activity.24 We also observed in our study that greater activation of the tibialis anterior is correlated with higher scores on the MIF scale.This means that the individuals who most activated this musculature also reached the best status on the functionality scale, which highlights the importance of this muscle for the activities of daily living.Another musculature worth mentioning is the vastus medialis.In the group of individuals without SCI, it was the most activated muscle in all modalities, mainly in the elliptical.The choice of a specific training device is influenced by therapeutic goals (e.g., strengthening muscles, improving reciprocal muscle activation, or simulating muscle activity patterns during walking) and the individual's functional abilities.25 The elliptical is an alternative that facilitates coordination between limbs; it involves reciprocal activity in an upright position, requiring the patient's body weight to be largely supported by their legs with the use of handlebars for balance support.The elliptical generates lower pedal reaction forces than ground reaction forces when walking, but greater Discussion The objective of this study was to verify the muscle activation pattern of the vastus medialis, gluteus medius, tibialis anterior and paraspinal muscles during exercises with elliptical trainer and ergometric bicycle with and without electromyographic biofeedback in individuals with ISCI.It was observed that in the group with ISCI, there was a greater activation of the tibialis anterior in the BG and BBG in relation to the EG and EBG.Furthermore, the use of biofeedback does not seem to interfere with muscle activation in the two modalities studied.In relation to the reference group, it is observed that the vastus medialis was the most activated musculature in all modalities, mainly in the elliptical trainer.However, on the bicycle, both the tibialis anterior and vastus medialis muscles showed significant activation.Individuals with ISCI have an important deficit in muscle activation below the level of injury, which significantly compromises many functional actions, such as gait, 6 clearly observed in this study and already well described in the literature.However, when submitted to specific activities, some musculatures stand out in relation to others.Both the subjects with ISCI and those in the RG achieved greater activation of the tibialis anterior muscle on the cycling exercise compared to the other modalities.
independence.ET_VM = activation of the vastus medialis in the elliptical trainer; ET_GM = activation of the gluteus medius in the elliptical trainer; ET_PS = activation of the paraspinal muscles in the elliptical trainer; ET_TA = activation of the tibialis anterior in the elliptical trainer; EB_VM = activation of the vastus medialis in the ergometric bicycle; EB_GM = activation of the gluteus medius on the ergometric bicycle; EB_PS = activation of the paraspinal muscles on the ergometric bicycle; EB_TA = activation of the tibialis anterior on the ergometric bicycle.Fisioter Mov.2023;36:e36131 9 bicycle and the elliptical trainer are effective to promote activation in the muscles studied, highlighting the greater activation of the vastus medialis in the elliptical modality, and the greater activation of the tibialis anterior in the bicycle.Being able to prescribe kinesiotherapy with reliable evidence on the activation of target muscles during an exercise makes clinical practice more objective and effective.In this sense, it is necessary new studies to better elucidate the activation of the muscles involved in gait and trunk control in the use of the equipment used in our research.Authors' contributionsPPS: Conception, data analysis and interpretation, and writing of the manuscript.MGG: Conception and, along with DH and TSR, obtaining data and preparing the manuscript.FZ and JSBA: analysis and interpretation of data and critical review of the manuscript.FC: Important intellectual criticisms of the content, final approval of the version to be published.knee and hip movement.26 Damiano et al. 23 compared muscle activity patterns of the quadriceps and hamstring muscles during four conditions: floor walking, treadmill walking, stationary cycling, and elliptical training.The authors observed greater activation of the quadriceps muscles on the elliptical trainer compared to other modalities, corroborating our findings.Burnfield et al. 22 add that the range of motion caused by the elliptical lifts the leg without weight vertically in a greater amplitude than walking, where the limb barely touches the surface.This causes a greater excursion in several planes, mainly in more proximal joints, demanding more proximal musculatures as well, which corroborates our result of greater activation of the vastus medialis on the elliptical in the RG.In addition, the bicycle also provided great activation in this musculature, by causing activation of muscles such as vastus intermedius, vastus lateralis, vastus medialis and rectus femoris in the knee extension during descending phase.21 There is evidence that neural circuits are shared in cycling and walking and that both require reciprocal motor coordination, 23 encouraging the choice of this equipment in gait rehabilitation.Regarding biofeedback, surprisingly no difference was found in its use in activating the tested muscles.Biofeedback is a technique to support motor learning and has been used in clinical practice.27 EMG therapy plus biofeedback is based on the improvement of myoelectric signals obtained from the muscles that have been converted into visual signals, with the aim of informing the subject about the activity of the muscles.28 Previous studies express similar results to our research.A systematic review investigated the effects of EMG with biofeedback on the recovery of motor function after stroke, also demonstrating the little benefit for motor recovery when this tool is added to other rehabilitation strategies.29 On the other hand, more current research demonstrates the benefit of biofeedback when it is applied over a longer period and with significant regularity.Dost Sürücü and Tezen, 30 for example, demonstrated improvements in range of motion, muscle strength, muscle tone, and functionality in subjects rehabilitated with EMG plus lower extremity biofeedback combined with conventional physiotherapy compared to those who were treated with conventional physiotherapy alone.However, the techniques involving this resource were performed for three weeks with a frequency of five days a week.30

Table 4
activation of the vastus medialis compared to the other two modalities (p= 0.002).Regarding the paraspinal muscles, no statistical difference was observed between the modalities.No difference was found with the inclusion of biofeedback in any exercise modality in either group.FISIOTERAPIA EM MOVIMENTO Physical Therapy in Movement Santos PP et al.Fisioter Mov.2023;36:e36131

Table 1 -
Sample characterization Student's t test and Fisher's exact test were performed for age data and right side evaluated, respectively.ISCIG = incomplete spinal cord injury group; RG = reference group; ASIA = American Spinal Injury Association Scale; FIM = Functional Independence Measure Scale.

Table 3 -
Percentage of the absolute electromyography value of the spinal injured group reached from the mean normalized by the reference group for each muscle

Table 4 -
Absolute values of muscle activation measured through electromyography in all exercise modalities Note: ISCIG = incomplete spinal cord injury group; RG = reference group; EG = elliptical group; EBG = elliptical + biofeedback group; BG = bicycle group; BBG = bicycle + biofeedback group; VM = vastus medialis; GM = gluteus medius; PS = paraspinal muscles; TA = tibialis anterior; Statistical analysis by median and interquartile range (IQR), presented with a significance level, with *p < 0.05 greater activation of TA in BG and BBG in relation to the other groups (only for the GLM); # p < 0.05 VM muscle in relation to other muscles (only in RG); & p < 0.05 VM and TA muscle in relation to other muscles (only in RG); @ p = 0.002 of the VM muscle of the EG and EBG in relation to the other two groups (only in the RG).Friedman test.

Table 5 -
Correlation of the Functional Independence Measurement Scale (FIM) score in relation to the muscles evaluated in the bike and elliptical modalities in the injured group Fisioter Mov.2023;36:e36131 8