You are viewing the site in preview mode

Skip to main content
Download PDF
Download ePub

Activity-based recovery training with spinal cord epidural stimulation improves standing performance in cervical spinal cord injury

Journal of NeuroEngineering and Rehabilitation volume 22, Article number: 101 (2025) Cite this article

Abstract

Background

Individuals with a clinically complete spinal cord injury are unable to stand independently without external assistance. Studies have shown the combination of spinal cord epidural stimulation (scES) targeted for standing with activity-based recovery training (ABRT) can promote independence of standing in individuals with spinal cord injury. This cohort study aimed to assess the effects of stand-ABRT with scES in individuals with cervical chronic spinal cord injury. We evaluated the ability of these individuals to stand independently from physical assistance across multiple sessions.

Methods

Thirty individuals participated in this study, all unable to stand independently at the start of the intervention. Individuals were participating in a randomized clinical trial and received stand-ABRT in addition to targeted cardiovascular scES or voluntary scES. During the standing intervention, participants were asked to stand 2 h a day, 5 days a week for 80 sessions (Groups 1 and 2) or 160 sessions (Groups 3 and 4).

Results

A total of 3,524 training days were considered for analysis. Group 1 had 507 days, group 2 with 578 days, and 1152 and 1269 days for groups 3 and 4 respectively. 71% of sessions reached the two-hour standing goal. All individuals achieved outcomes of lower limb independent extension with spinal cord epidural stimulation, with a wide range throughout a training day. Sixteen participants achieved unassisted hip extension while maintaining unassisted bilateral knee and trunk extension. Participants receiving initial voluntary scES training performed better in unassisted bilateral knee and trunk extension than those receiving initial cardiovascular scES. The lower-limb standing activation pattern changes were consistent with the greater standing independence observed by all groups.

Conclusions

Individuals with chronic cervical spinal cord injury were able to achieve various levels of extension without manual assistance during standing with balance assist following stand-ABRT with scES. These results provide evidence that scES modulates network excitability of the injured spinal cord to allow for the integration of afferent and supraspinal descending input to promote standing in individuals with spinal cord injury.

Trial registration

The study was registered on Clinical Trials.gov (NCT03364660) prior to subject enrollment.

Background

Spinal cord injury (SCI) can result in paralysis and autonomic dysregulation among other consequences. In general, individuals with a severe cervical SCI cannot stand independently without external assistance. The loss in mechanical loading due to paralysis has secondary consequences such as rapid loss of skeletal muscle [1, 2]. In turn, this loss of muscle has been linked to an increase in intramuscular fat which predisposes individuals to an increase risk of type-2 diabetes and cardiovascular disease [3, 4]. Functional electrical stimulation and neuromuscular electrical stimulation of the lower extremities have been used to mitigate the reduction in skeletal muscle and improve overall health [5]. Similarly, interventions focused on exercise and loading are critical for overall health in individuals with SCI [6,7,8]. In particular standing has been identified by individuals with SCI to maintain or improve health [7].

Activity-based recovery training (ABRT) reinforces spinal circuits through the activation of sensory and motor pathways and can improve performance of practiced tasks [9, 10]. The importance of weight-bearing training has been supported over the years with many studies focused on improving locomotion following SCI [9, 11, 12]. However, there is a lack of focus on standing interventions for individuals with motor complete paralysis. Standing ability is typically synonymous with balance as quantified through scales such as the Berg Balance Scale, which assesses static and dynamic balance elements [13,14,15]. Two recent systematic reviews evaluated the effectiveness of various training interventions on improving balance in individuals with incomplete SCI, revealing either marginal or no improvements with interventions such as locomotor training, robotic training and functional electrical stimulation [16, 17]. This phenomenon can be explained by the lack of specificity of the interventions received by the study population. The need to study and quantify the benefits of standing interventions has long been recognized [18], however there is a lack of reports on the effectiveness of standing interventions for recovery of full weight-bearing standing in motor complete populations.

Lumbosacral spinal cord epidural stimulation (scES) has been a successful intervention for recovery of function in individuals with SCI [19]. Proof of principles studies have shown the combination of stimulation targeted for standing (Stand-scES) with ABRT promote independence of standing with balance assist in the SCI population [20,21,22]. Prior studies have shown that the spinal circuitry can interpret sensory information related to loading to promote effective lower-extremity extension patterns in the presence of Stand-scES [20, 21, 23]. This study expands on previous reports of standing ability by identifying the progression of unassisted standing in individuals with cervical chronic SCI throughout a Stand-ABRT intervention with Stand-scES. In addition, we quantify EMG changes before and after intervention. We hypothesized that individuals with severe cervical SCI would improve on their ability to stand as determined by time with unassisted bilateral knee extension throughout a Stand-ABRT intervention.

Methods

Clinical characteristics of study participants

We enrolled 30 individuals with chronic cervical SCI. The average age was 36.8 ± 10.4 years with an average time since injury of 11.8 ± 9.2 years. Sixteen were classified as AIS-A, and 13 were female (Table 1). All participants were unable to stand physically unassisted at the start of the intervention. All individuals were participating in a randomized trial to evaluate the efficacy of cardiovascular targeted scES (CV-scES) on blood pressure regulation. The standing intervention was added to assess the effect of weight-bearing on the primary outcome related to cardiovascular regulation. In the current paper we only report standing related outcomes during the standing intervention, however all participants were concurrently receiving an additional modality of stimulation (see Training Intervention section). All individuals provided written informed consent approved by the University of Louisville’s Institutional Review Board. The study was registered in clinical trials.gov before participant enrollment (NCT03364660). Participants were recruited from 2019 to 2022. All participants were implanted with a 16-electrode array (5-6-5 Specify; Medtronic, Minneapolis, MN, USA) placed between spinal segments L1-S1 and an Intellis™ neurostimulator (Medtronic, Minneapolis, MN, USA ) as previously described [24]. Neurophysilogical monitoring during the surgical implantation was used to guide the placement of the electrode array [24].

Table 1 Clinical characteristics
Full size table

Stimulation configurations

Before the start of interventions, participants were mapped for all possible interventions (CV-scES, Vol-scES and Stand-scES) as previously described [24]. Stimulation parameters were task-specific and participant specific. More explicitly, for standing, we used electromyography (EMG) of the lower extremities and trunk to evaluate motor activation during scES. We adjusted stimulation parameters to modulate extensors during weight-bearing and achieve unassisted ankle, knee and hip extension during standing. As previously described, cathodes and anodes were selected to promote lower limb extension, amplitude and frequency were subsequently adjusted to sufficiently activate extensors without generating a bursting pattern [24]. At the conclusion of all mapping sessions, participants were randomized into a study group (Fig. 1).

Stand-scES parameters were adjusted throughout the standing intervention for each participant. These adjustments were performed to promote the greatest level of unassisted standing as possible. First and last stimulation parameters for all participants are presented in the supplementary materials (Additional file 1).

Training interventions

As part of the randomized study design, all groups received at least one intervention that included Stand-ABRT in combination with scES targeted for standing. Following the implantation of the neurostimulator, participants were randomized to one of four study groups (Fig. 1). All groups received targeted cardiovascular scES (CV-scES) or voluntary movement scES (Vol-scES) as part of their daily training. Groups 1 and 2 performed CV-scES or Vol-scES, respectively, for 80 sessions (Intervention 1) with no Stand-ABRT. During intervention 2, these groups continued their CV-scES or Vol-scES and had an additional 2-hours of Stand-ABRT. Groups 3 and 4 performed the same combination of CV-scES or Vol-scES plus the additional 2-hours of Stand-ABRT for both interventions 1 and 2. Stand-ABRT always occurred with Stand-scES in the laboratory, for either 80 visits (n = 15: Groups 1 and 2) or 160 visits (n = 15: Groups 3 and 4).

Fig. 1
figure 1

Randomized clinical trial design. Following implantation individuals were randomized into one of four groups. Group 1 received CV-scES for 80 sessions followed by 80 sessions of CV-scES and Stand-ABRT with Stand-scES. Group 2 received Vol-scES for 80 sessions followed by 80 sessions of Vol-scES and Stand-ABRT with Stand-scES. Group 3 received CV-scES and Stand-ABRT with Stand-scES for 80 sessions and repeated the same intervention for an additional 80 sessions. Group 4 received Vol-scES and Stand-ABRT with Stand-scES for 80 sessions and repeated the same intervention for an additional 80 sessions

Full size image

During the stand ABRT intervention phase, participants were asked to stand with Stand-scES for 2-hours each day, 5 days a week, with the least amount of manual assistance as possible. Participants were allowed to take sitting breaks as necessary. Sitting time was not counted towards their 2-hour target. Two to four research staff trained in activity-based interventions provided assistance-as-needed to maintain the joint/body segment in proper kinematics and posture. Participants achieved standing independence when no external assistance was required to maintain the joint in proper kinematic position with good standing posture. In all cases, bouts of independence still required close monitoring by research staff ready to provide assist as needed. We used a custom LabVIEW program (National Instruments, Austin TX USA) to record assistance levels and duration for each joint (hip, knees and ankles) during all Stand-ABRT sessions (Additional File 2, Fig. S1). In addition, we monitored levels of assistance for the trunk as well as upper extremity self-assist modality. The research staff assigned to the ABRT session verbalized ‘independence’ at a specific joint when no assistance was required, and ‘back on’ when assistance was again required. Another team member would record these changes in assistance levels in the custom program. Stand-ABRT sessions could occur on a body weight support (BWS) system (n = 6), a custom standing frame (n = 19), or a fixed walker (n = 5). The custom standing frame has been previously described and did not constrain the participant’s lower extremities or trunk, allowing 100% weight-bearing [21]. The participant could use the frame structure for upper extremity support and balance assist. The fixed walker provided less support compared to the standing frame. The fixed walker consisted of fixing a standard walker to a base to prevent tipping. Participants who began training on a BWS or a standing frame could be transitioned to a different device during the ABRT intervention period.

Experimental sessions

Standing activation pattern with Stand-scES was assessed during a dedicated experimental session per time point, that is at pre-intervention (prior to randomization), post-intervention 1 and post-intervention 2. In these assessments, the goal for the participant was to stand for 30 min with the least amount of external assistance provided manually by trainers. Seated resting periods occurred when requested by the individual. Standing activation pattern was assessed for the longest standing bout performed with the least amount of external assistance, which needed to be achieved for a minimum duration of 1 min in order to be considered for this analysis. All assessments were performed in a standing frame or BWS for those unable to stand overground. Force plates (Kistler, Holding AG, Winterthur, Switzerland) and EMG data were collected at 2000 Hz by a hard-wired AD board and custom-written acquisition software (LabView, National Instruments, Austin, TX). Force data were low-pass filtered (10 Hz) and EMG data band-pass filtered (10–500 Hz). EMG was collected bilaterally from the gluteus maximus, rectus femoris, vastus lateralis, medial hamstrings, tibialis anterior, medial gastrocnemius, and soleus. Time- and frequency-domain EMG were averaged across all muscles within each participant and time point and considered for analysis [22, 25]. In particular, EMG pattern variability (EMG CV) was quantified by assessing the coefficient of variation of the EMG linear envelope [21]. EMG total power was normalized by the maximum value detected within each participant across time points, and continuous Wavelet Transform was applied to calculate EMG median frequency (EMG MDF) and its standard deviation (EMG MDF SD) [25]. Previous analysis of a large standing dataset from our group showed that the proposed EMG parameters, averaged across key lower limb muscles, were significantly different between knee-assisted and knee-independent standing, with independent extension favored by higher EMG total power as well as lower EMG CV, EMG MDF and EMG MDF SD [25]. Finally, the coefficient of variation of the vertical ground reaction force (GRF CV) was also considered for analysis. No EMG data is reported for participant C237.

Outcome measures

We used a custom Matlab program (MathWorks, Carlsbad, CA, USA) to automatically calculate the total time (in seconds) for multiple combinations of independence variables (Table 2).

Table 2 Extracted variables for analysis
Full size table

The following EMG outcomes were extracted for analysis EMG CV, EMG MDF, EMG MDF SD, GRF CV. All EMG outcomes were compared at the Pre-Intervention, Post Intervention 1 and Post Intervention 2 time points.

Statistical analysis

Total standing time was calculated based on weight-bearing minutes for each participant across all training days. An intent-to-treat approach was used for data analysis, a training occurrence was included in the analysis if the day contained weight-bearing minutes. One participant in group 4, withdrew after completion of the first intervention. Two participants in group 1 withdrew after 25 and 26 sessions respectively, their data for all available sessions up until withdrawal was included in the analysis.

Standing outcomes were summarized with median ± median absolute deviation (MAD) since more than half were not normally distributed per the Shapiro-Wilk test. MAD represents the median distance from any value to the median. Within-group changes were evaluated with signed Rank test and between-groups differences of changes were evaluated with the Wilcoxon Rank Sum test.

To quantify the comparisons estimates, a non-parametric equivalence to Cohen’s d [26] effect size (ES) was calculated as median divided by MAD (ES = median/MAD). Sawilowsky’s extension to Cohen’s criteria was used to classify the obtained effect sizes as trivial (< 0.01), very small (0.01–0.2), small (0.2–0.49), medium (0.5–0.79), large (0.8–1.19), very large (1.2–1.99), and huge (> 2.0). The threshold of 0.5, found to be the minimally important difference in health-related quality of life was used for meaningfulness [27, 28].

To quantify group comparisons, standing activation pattern values were evaluated with mixed linear models with random intercept included per participant to account for individual variability, group, timepoint (Pre/Post1/Post2 or Pre/Post), as well as their interaction were included as factors in those models. Least square means, and standard error from the models were used to represent the data. Different comparisons, change over time, difference between groups at specific timepoints, and difference of changes (interaction) were performed through linear contrasts built on the interaction term.

For all tests and models, the significance level was set to 0.05. All tests were 2-sided. The statistical analysis was performed in SAS 9.4.

Results

A total of 3,524 training days were considered for analysis. 70% of these training days reached the 2-hour weight-bearing goal, while 93% surpassed 1-hour standing. There was no difference in the number of sessions that did not reach two hours, when comparing the initial 80 training sessions to the second 80 sessions, for those in groups 3 and 4. Reasons that individuals did not reach the 2-hour standing target included: fatigue, orthostatic hypotension, bladder or bowel accidents, or logistical reasons such as participant’s late arrival to the laboratory (Fig. 2). Fatigue was the primary reason reported in 40% of the occurrences (Fig. 2A). Participants running out of time during their assigned training block was reported in 16.5% of the occurrences. This was typically due to multiple requests for long rest periods during the session. Other logistical reasons (1.3%) were due to participants’ late arrival resulting in the inability to complete the session within their allotted time block. Individual data distribution for all available sessions is presented in Fig. 2B.

No serious adverse events took place during the intervention period. A summary of the adverse events experienced while participants were in an active Stand-ABRT intervention are presented in the supplementary materials (Additional File 3). A total of 66 related and unrelated adverse events were reported.

Fig. 2
figure 2

Incidence of inability to reach two-hour weight-bearing target. (A) Reported or measured rationale for ending the standing session prior to 2-hr target. (B) Box Plots including individual data for sessions that did not reach the 2-hr target

Full size image

Stand-scES parameters were adjusted throughout the Stand-ABRT intervention to respond to plasticity and promote better standing ability. Within a training session, amplitude could be adjusted, typically within a range of 0.1 to 0.5 mA, most typically taking place later in the session due to loss of extension as a result of muscle fatigue. Configuration changes adjusting cathodes or anodes were performed less frequently. Participants in group 1 had on average 3.4 configurations (+/- 1.8), similarly participants in group 2 had on average 3.1 configurations (+/- 1.7). Those in groups 3 and 4 had larger number of configurations 5.9 +/-2.9 and 5.5 +/- 3.0 respectively. The larger number of used configurations can be suggested to be the result of two training interventions as opposed to a single standing intervention.

Trunk independence

Twenty-seven of 30 individuals reached some ability to maintain appropriate trunk posture without external assistance. Three individuals performed all sessions in the body-weight-support, requiring weight-support to maintain their trunk in proper posture. Across all participants, the median trunk independence was 63.8% of stand-time for the initial 80 training days, and 94% of stand-time for the second set of 80 training days (Fig. 3A). An important observation revealed that trunk independence duration does not depend on neurological level of injury (Fig. 3B). A correlation of r = 0.35 was found between level of injury and median trunk independence. Participants with injury levels C4-C8 were able to achieve a median of 120-minutes of trunk independence across all sessions. Trunk independence duration tended to be consistent across sessions, particularly on those individuals that were able to achieve full trunk independence early in the intervention (Fig. 3C). Participants that received an additional 80 sessions (Groups 3 and 4) showed similar trends with higher median values (Fig. 3D). Trunk independence was a factor affecting the decision for assistive device progression, as a result 17 participants progressed to a less assistive device, including two participants that progressed from the body-weight-support system to a fixed walker.

Fig. 3
figure 3

Trunk Independence Levels during Standing. (A) Percent of total Standing time across all sessions with trunk independence for the first and second standing intervention. Box plots are mean +/- 1 standard deviation, lines represent 95% confidence interval. Median line is red. (B) Radial graph showing relationship between neurological level of injury and median trunk independence minutes across sessions. Inter-circles are 20 min intervals from 0–120 min. (C) Color map (left) shows the trunk independence time during each training session (rows) across participants (columns) for the two-hour intervention during their initial standing intervention (80 sessions). Top row indicates the first session. Box plot (right) representing trunk independence time for each participant. Median is represented in red. (D) Color map (left) shows the trunk independence time during each training session (rows) across participants (columns) for the two-hour intervention during their second standing intervention (Groups 3 and 4). Top row indicates the first session of the second standing intervention. Box plot (right) representing trunk independence time for each participant. Median is represented in red

Full size image

Knee independence

All 30 individuals were able to achieve outcomes of lower limb unassisted extension, ranging from 10 min of unassisted single knee to 118 min of unassisted bilateral knees within a training day. Seventeen of 30 participants had at least 50% of their total stand-time across all sessions with one or both knees independent during the first 80 training days, including seven participants that were able to maintain knee independence for over 80% of total stand-time (Fig. 4A). Additionally, 11 of 14 participants that completed a second set of 80 training days (Groups 3 and 4) had at least 50% of their total stand-time with one or both knees independent (Fig. 4B).

Fig. 4
figure 4

Lower-extremity Independence Levels during Standing. (A) Percent of total Standing time across all sessions with left knee (gray), right knee (blue) or bilateral (pink) independence. Each line is a participant during their initial standing intervention (80 sessions). (B) Percent of total Standing time across all sessions with left knee (gray), right knee (blue) or bilateral (pink) independence. Each line is a participant during their second standing intervention (Groups 3 and 4). (C) EMG features in time and frequency domains collected during standing pre and post intervention. EMG features values were averaged among participants and among all assessed muscles. Values are expressed as median + mean absolute deviation. Differences were tested by Wilcoxon test * p < 0.05. (D) EMG features in time and frequency domains collected during standing post intervention 1 and post intervention 2 for groups 3 and 4. EMG features values were averaged among participants and among all assessed muscles. Values are expressed as median + mean absolute deviation. (E) Color map (top) shows the bilateral independence time during each training session (rows) across participants (columns) for the two-hour intervention during their initial standing intervention (80 sessions). Top row indicates the first session. Box plot (bottom) representing bilateral independence time for each participant. Median is represented in red. (F) Color map (top) shows the bilateral independence time during each training session (rows) across participants (columns) for the two-hour intervention during their second standing intervention (Groups 3 and 4). Top row indicates the first session of the second standing intervention. Box plot (bottom) representing bilateral independence time for each participant. Median is represented in red

Full size image

Improvements in knees independence during the first 80 stand training days (Fig. 4A and C) were associated with lower EMG CV, 0.35 + 0.22 vs. 0.28 + 0.14; (p = 0.048) observed after Stand-ABRT (Fig. 4C). Conversely, no trends were observed when comparing standing activation patterns generated after 80 and after 160 stand training days by the subgroup of participants (Groups 3 and 4) that underwent the second 80 sessions of Stand-ABRT (Fig. 4D).

The progression of acquiring bilateral knee independence varied across participants. Some individuals were able to reach bilateral independence for two-hours as early as session 20 (Fig. 4E). However, all participants demonstrated substantial variability in the number of independent minutes per session. Three individuals from groups 3 and 4 had reduced variability during intervention 2 (Fig. 4F). There was a tendency towards greater variability (MAD +/- 28.5 to 39.4 min) in the middle range (45–90 min of unassisted bilateral knees), while those that had lower independent times were consistently low (MAD +/- 8.0–0.0 min). Figure 5 shows two representative standing patterns collected from the same research participant prior to any training (left, Pre-Intervention) and after stand training with Stand-scES and CV-scES (right, Post-Intervention 1). The activation pattern generated Pre-Intervention was not effective to achieve independent standing, and consisted in the alternation between EMG bursts and negligible activity (i.e. similar to a rhythmic pattern) of lower limb muscles. After stand training with Stand-scES and CV-scES, the same participant demonstrated an overall continuous (i.e. non-rhythmic) EMG pattern with lower level of activation of primary lower limb muscles which resulted in the ability to achieve bilateral independent extension. When assessing overall knee independence and comparing it to independence of knees and trunk, time with combined knee and trunk independence was lower during the initial 80 sessions of Stand-ABRT (see Additional File 2, Fig. S2 and Additional File 4). During the second intervention of Stand-ABRT (Groups 3 and 4), most participants could account for all their knee independence time with trunk independence. Only 5 participants had unassisted knee extension that was present during trunk assistance (see Additional Files 2, Fig. S2).

Fig. 5
figure 5

Representative standing activation patterns pre and post intervention. EMG collected from participant A123 (left lower limb) prior to any intervention (Pre-Intervention) and after 80 sessions of stand training with epidural stimulation (Stand-scES) combined with epidural stimulation for cardiovascular function (Group 3; Post-Intervention 1). The participant required manual assistance at both knees Pre-Intervention, while independent knees extension was achieved Post-Intervention 1. Trunk and hips were manually assisted for all standing bouts reported. The two representative activation patterns per time point were collected within the longest standing bout considered for analysis, 4 min apart at Pre-Intervention and 8 min apart at Post-Intervention 1. Baseline EMG collected during sitting without scES is also shown (to the left of standing panels). Note the more variable EMG and vertical ground reaction force (Force) pattern at Pre-intervention when manual assistance at the knees was required. GL, gluteus maximus; MH, medial hamstring; RF: rectus femoris; VL, vastus lateralis; TA, tibialis anterior; SOL, soleus

Full size image

Full standing independence

Seventeen participants achieved hip independence while maintaining bilateral knee and trunk independence (Fig. 6A). An additional movie shows this in more detail (see Additional File 5). This fully independent standing ranged from less than a minute (n = 2) to 107 h throughout all ABRT sessions. Figure 6B shows the maximum total unassisted time achieved in a single session for those that demonstrated the ability to stand without manual assistance. Individuals in group 1 achieved the lowest levels of full independence, while individuals in groups 2 and 3 had sessions with the greatest independence (Fig. 6B and C). The participant achieving the greatest level of independence had episodes of full independence in 82% of her sessions with a range of 22 s to 137 min with a median of 103 min and MAD +/-14.85 (Fig. 5C; A110).

Fig. 6
figure 6

Standing without manual assistance. (A) Photograph of individual (B41) standing with independent trunk, hips and bilateral knees using a walker for balance assist. (B) Maximum independent minutes standing with independent trunk, hips and bilateral knees during single session. Each data point is a participant (n = 16). Individuals above the 40-minute line are shown in plot C (n = 8). (C) Box plot showing session time standing with independent trunk, hips and bilateral knees for individual participants. Group 2 (Vol-scES): blue dashed, Group 3 (CV-scES + Stand-scES): red fill, Group 4 (Vol-scES + Stand-scES): blue fill. Box plots are mean +/- 1 standard deviation, lines represent 95% confidence interval

Full size image

Standing ability and activation pattern - group comparisons

Standing activation pattern generated prior to any intervention was not statistically different across the four groups, with the sole exception of EMG MDF between groups 2 and 3 (p = 0.039). When comparing lower extremity standing independence across groups there is a very large effect size of 1.35 (p = 0.47) for greater bilateral knee independence in those participants that first received Vol-scES during intervention 1 (Group 2) as compared to those participants using CV-scES (Group 1) (Fig. 7A). The larger standing independence shown by group 2 was associated with activation pattern adaptations consisting of lower EMG CV (p = 0.023) and lower EMG MDF SD (p = 0.028). Conversely, no standing activation pattern trends were observed in group 1 (Fig. 7A). The group that received an initial 80 sessions of CV-scES alone (Group 1) had lower bilateral knee independence when compared to the group receiving CV-scES + Stand-scES (Group 3) (Fig. 7B). This difference was quantified with a large effect size (Cohen’s d = 1.01; p = 0.31).

Fig. 7
figure 7

Group comparisons of knee independence during standing. (A) Top: left, right and bilateral knee independence comparison for Group 1 (CV) and Group 2 (Voluntary). Bottom: EMG features in time and frequency domains collected during standing post intervention 1 vs. post intervention 2. Values were averaged among participants and among all assessed muscles. Values are expressed as median + MAD. (B) Top: left, right and bilateral knee independence comparison for Group 1 (CV) and Group 3 (CV-Stand). The graph compared the initial 80 sessions of standing (intervention 2 for group 1; and intervention 1 for group 3). Bottom: EMG features in time and frequency domains collected during standing pre vs. post intervention 1 for Group 3 and post intervention 1 vs. post intervention 2 for Group (1) (C) Top: left, right and bilateral knee independence comparison for Group 2 (Voluntary) and Group 4 (Voluntary-Stand). The graph compared the initial 80 sessions of standing (intervention 2 for group 2; and intervention 1 for group 4). Bottom: EMG features in time and frequency domains collected during standing pre vs. post intervention 1 for Group 4 and post intervention 1 vs. post intervention 2 for Group (2) Differences assessed by Wilcoxon test * p < 0.05. (D) Top: left, right and bilateral knee independence comparison between intervention 1 and intervention 2 for Group 3 (CV-stand). Bottom: EMG features in time and frequency domains collected during standing pre vs. post intervention 1 vs. post intervention 2. (E) Top: left, right and bilateral knee independence comparison between intervention 1 and intervention 2 for Group 4 (Voluntary-stand). Bottom: EMG features in time and frequency domains collected during standing pre vs. post intervention 1 vs. post intervention 2. (F) Box plot comparing trunk independence across groups and interventions. Box plots are mean +/- 1 standard deviation, lines represent 95% confidence interval

Full size image

In contrast, group 2 that received 80 sessions of Vol-scES prior to standing performed better for bilateral independence when compared to group 4 (Vol-scES + Stand-scES) (Fig. 7C). However, this difference (p = 0.60) had a mediun effect size (Cohen’s d:0.74). Groups 3 and 4 both improved their ability to stand with independent bilateral knees during intervention 2 when compared to the initial 80 sessions (Fig. 7D and E). Effect sizes for total bilateral independence time change between intervention 2 vs. intervention 1 were large for groups 3 and 4 (p = 0.22, Cohen’s d:0.81 and p = 0.03, d = 0.87 respectively). No standing activation pattern trends were observed between intervention 2 and intervention 1 in these two groups, except for the larger EMG CV (p = 0.008) shown by group 4 (Fig. 7E). Trunk independence time was greater in the voluntary groups (Groups 2 and 4) when compared to the CV groups (Groups 1 and 3) (Fig. 7F). The effect size difference between groups 1 and 2 was 0.72 (p = 0.85). Similarly to knee independence, the group that received 80 sessions of Vol-scES prior to Stand-ABRT (Group 2), had higher trunk independence values than group 4 at their initial 80 sessions of Stand-ABRT (p = 0.16)(Fig. 7F).

Discussion

This study describes the progression of standing ability with Stand-scES during an ABRT protocol with scES for individuals with severe cervical SCI. Two hours of standing was achieved by 97% of participants by the conclusion of the intervention period. With task-specific tonic continuous stimulation, all participants could stand with at least a single leg unassisted, and 26 could demonstrate episodes of unassisted bilateral knee extension. Remarkably, 16 participants achieved episodes of simultaneous unassisted hip, knee and trunk extension while using upper extremity for balance assistance.

The ability to complete a two-hour intervention varied greatly by participant. The most often provided reason for ending the session early was fatigue. Lack of adequate physical activity before trial enrollment could explain how fatigue influenced an individual’s ability to complete a two-hour standing intervention [29,30,31]. A total of 10 participants had sessions end early as a result of fatigue during the first intervention. While 7 participants had sessions end early during the second intervention as a result of fatigue. One participant (B213) never achieved 2 h throughout both her interventions (Group 4). While another participant (A232) did not achieve 2 h during the first intervention but was able to reach 2 h in 91% of her sessions in the second intervention (Group 3). Group 4 had the greatest report of fatigue resulting in days with less than 2 h, 116 days reported by 5 out of 8 participants (mean: 23.2 +/- 33.1 days). Group 2 had the lowest number of reported days, 14, all reported by one participant (C192). Group 1 and 3 had two participants each with means of 32.5+/-38.9 days and 41.4+/-55.9 days respectively. We explored potential links between the number of sessions not meeting the 2-hour duration as a result of fatigue and factors such as age and time since injury. Pearson correlation r values were 0.24 and − 0.08 for age and time since injury respectively. A slightly stronger correlation was found between reported fatigue days and stimulation amplitude (Pearson’s r = 0.42). Although we did not actively explore how each individual defined fatigue and cannot therefore discard a component of neuromuscular fatigue, we presume that neuromuscular fatigue could be linked to loss of extension resulting in assisted standing rather than early termination of the session. An association could be argued with stimulation intensity driving higher muscle contractions and increasing the overall sense of feeling tired. However, in general, a trend was not found between shorter sessions resulting from fatigue occurring earlier in the intervention period. Although cardiovascular dysregulation is typical in this population [32,33,34], the ability to interleave stimulation cohorts that assisted in maintaining blood pressure stable during standing greatly reduced the incidences of hypotension during standing [24]. The low percentage of sessions that had to end prematurely due to hypotension (0.5% of total standing sessions) could be attributed to the capability of scES to regulate blood pressure while also targeting appropriate extensor activation as was previously shown by Aslan et al. [35].

Participants ranged in injury level from C3-C8, demonstrating various levels of trunk control at study enrollment. Median independent trunk time across 80 sessions of Stand-ABRT was 63.8% (IQR: 37.8–96.7), the time of trunk independence varied among participants as shown in Fig. 3. A uniqueness of this study was the broad characteristics of the enrolled group. Nonetheless, participants across all neurological levels could achieve standing without trunk assistance. This provides initial evidence that Stand-scES has the potential to access lower-extremity as well as trunk motor pools. The combination of scES and Stand ABRT provides sufficient excitation below the level of injury to promote increased neuromuscular activation at all levels of trunk and legs. There is prior evidence of improvements in trunk stability in the presence of scES during sitting, suggesting greater activation of trunk musculature providing postural control [36, 37]. Those participants not achieving trunk independence were never able to transition from the BWS to an overground setting. Of note is the need for upper extremities balance assistance while standing. We instructed participants to minimize weight transferred through the arms. However, this was only a measured outcome during the experimental assessments and intervention. Trunk perturbations were not part of the intervention, however, as previously demonstrated, motor activity increases in response to perturbations when hands are not used for balance support [38]. The same study also showed that trunk displacement increased by 116% when supported by hands-on. Using upper extremities for balance assist allows individuals to feel safer reaching and challenging trunk posture while standing. Some participants could stand without trunk support and upper extremity balance assist; however, they required hip assistance to maintain balance. An additional movie shows this in more detail (see Additional File 6). Control of posture and upright standing requires neuromechanical contributions [39, 40]. Within the context of the kinetic chain, biomechanical changes at one joint affect other joints. It has been demonstrated that trunk stabilization is a critical response to platform tilts, emphasizing the importance of proper trunk alignment during standing. Further neuromuscular control is required to maintain the center of pressure within the base of support, activating various multi-directional muscle synergies [41]. The ability to stand with unassisted trunk and knees in the presence of scES suggests an improvement in spinal postural control contributing to the integration of sensory feedback and supraspinal control to effectively activate postural musculature.

Given the lack of outcome measures available to assess standing ability in individuals with clinically motor complete SCI, we focused our analysis on the quantification of unassisted extension at the knee and hip joints, which was measured as a time variable. The ability to maintain unassisted extension is due to the appropriate activation of lower extremity muscles. Our results suggest that an ABRT intervention with task-specific scES promotes motor activation leading to the ability to maintain knee extension during full weight-bearing standing. Full weight-bearing standing with balance assist while using neuromodulation has been reported in prior studies [20, 21, 42,43,44], suggesting that the human spinal circuitry has the potential to integrate sensory information to generate appropriate motor patterns necessary to maintain standing. The altered excitability of the spinal circuits induced by scES promotes the modulation of motor output in response to the integrated efferent signals. Therefore, in its most effective cases, scES fosters some level of voluntary control over extension during the standing bouts, this is shown detail in the supplementary video (see Additional File 7). Stimulation parameters including location, frequency and amplitude are crucial drivers for spinal circuitry modulation and the generation of motor output [45, 46]. It has been shown that different stimulation frequencies at the same site of stimulation can elicit various motor patterns, providing convincing evidence for task specific scES parameters [20, 24, 46].The standing activation pattern trends generally associated with the larger standing independence improvements in this study are consistent with previous findings while comparing independent and assisted standing in a population of motor complete SCI individuals implanted with scES [25], suggesting that a lower limb activation pattern that is overall more continuous (rather than variable) can promote better standing ability in this population. The variability in stimulation paradigms received among groups, combined with the variability in standing performance and the potential for differences experienced during assessments compared to training days hinder the interpretation of the EMG outcome data. It is important to note that during EMG assessments participants stood with a different standing frame as compared to the stand training sessions in addition to EMG and motion capture instrumentation, in some instances these factors resulted in altered performance. Our current study demonstrates the progression of how improved standing ability was achieved following a standardized ABRT intervention using neuromodulation. Significant variability was observed in the ability to maintain bilateral knee extension; however, some individuals could reach two hours of unassisted knee extension for a large portion of the intervention (Fig. 4).

The study design offered an opportunity to perform group comparisons to investigate the effect of multi-modal stimulation combining Stand-scES with autonomic (CV-scES) or motor (Vol-scES) training. The comparison of groups 1 and 2 offer some insight on pretraining autonomic regulation and motor recovery as precursors to standing. Those that trained with Vol-scES, focused on intentional movement of legs and trunk in supine and sitting positions (Group 2), showed improved standing performance over those that trained with CV-scES (Group 1). The ability to integrate supraspinal control with scES to generate intentional movements translated to arguably increased control over extension during standing. On the other hand, The CV-scES parameters are focused on preventing muscle activation and promoted the regulation of blood pressure through the activation of sympathetic vasomotor efferents [47]. Although as shown by group 3, asynchronous training with CV-scES and Stand-scES supports positive outcomes relative to standing ability. However, pre-training with CV-scES does not boost standing performance, suggesting specificity of motor activation is required for improved performance.

One limitation of the study is that the original randomized clinical trial was designed to evaluate efficacy of targeted scES for cardiovascular function and not to assess standing ability in individuals with cervical SCI. The low number of participants in each group limited the statistical comparison of standing ability or standing muscle activation patterns across groups. Further work is needed to quantify the interaction between multiple targeted stimulation paradigms on motor recovery in individuals with SCI. Additional work is also required to determine how training duration and frequency affect recovery. Similarly, a study specifically designed to prospectively evaluate recovery of unassisted standing in individuals with motor complete spinal cord injury could help elucidate some of the factors characterizing low-performers compared to high-performers.

Conclusions

Our study shows the potential of Stand-ABRT combined with task specific scES to promote the ability of individuals with chronic cervical SCI to recover unassisted full-weight bearing standing. Although 28 of 30 individuals were classified as motor complete, the intervention focusing on neuroplasticity principles maximizing load and sensory cues, paired with scES was able to promote the integration of sensory signals to maximize recovery in most participants. These results provide evidence that Stand-scES modulates network excitability of the injured spinal cord to allow for the integration of afferent and supraspinal descending input to promote standing in individuals with SCI. Asynchronous training with CV-scES or Vol-scES does not prevent recovery of standing in individuals with SCI.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Abbreviations

SCI:

Spinal cord injury

scES:

Spinal cord epidural stimulation

ABRT:

Activity-based recovery training

CV-scES:

Cardiovascular scES

Vol-scES:

Voluntary movement scES

EMG:

Electromyography

BWS:

Body weight support

EMG CV:

EMG pattern variability

EMG MDF:

EMG median frequency

EMG MDF SD:

EMG median frequency standard deviation

GRF CV:

Coefficient of variation of the vertical ground reaction force

MAD:

Median absolute deviation

ES:

Effect size

References

  1. Singh R, Rohilla RK, Saini G, Kaur K. Longitudinal study of body composition in spinal cord injury patients. Indian J Orthop [Internet]. 2014;48(2):168–77. Available from: https://www.ncbi.nlm.nih.gov/pubmed/24741139

  2. Castro MJ, Apple DF Jr., Staron RS, Campos GE, Dudley GA. Influence of complete spinal cord injury on skeletal muscle within 6 mo of injury. JApplPhysiol. 1999;86(1):350–8.

    CAS  Google Scholar 

  3. Cragg JJ, Noonan VK, Krassioukov A, Borisoff J. Cardiovascular disease and spinal cord injury: results from a national population health survey. Neurology. 2013/07/26. 2013;81(8):723–8.

  4. Cragg JJ, Noonan VK, Dvorak M, Krassioukov A, Mancini GB, Borisoff JF. Spinal cord injury and type 2 diabetes: Results from a population health survey. Neurology [Internet]. 2013; Available from: http://www.ncbi.nlm.nih.gov/pubmed/24153440

  5. Gorgey AS, Khalil RE, Carter W, Ballance B, Gill R, Khan R et al. Effects of two different paradigms of electrical stimulation exercise on cardio-metabolic risk factors after spinal cord injury. A randomized clinical trial. Front Neurol. 2023;14(September).

  6. Hicks AL, Martin KA, Ditor DS, Latimer AE, Craven C, Bugaresti J et al. Long-term exercise training in persons with spinal cord injury: effects on strength, arm ergometry performance and psychological well-being. Spinal Cord [Internet]. 2003;41(1):34–43. Available from: http://0.190.165.239.

  7. Eng JJ, Levins SM, Townson AF, Mah-Jones D, Bremner J, Huston G. Use of prolonged standing for individuals with spinal cord injuries. PhysTher. 2001;81:1392–9.

    CAS  Google Scholar 

  8. Forrest GF, Sisto SA, Harkema S, Kirshblum S, Wilen J, Bond Q. The effects of locomotor training on muscle activation, body composition, and bone mineral density. JSCM. 2005;28:465–6.

    Google Scholar 

  9. Dietz V, Harkema SJ. Locomotor activity in spinal cord-injured persons. JApplPhysiol. 2004;96(5):1954–60.

    CAS  Google Scholar 

  10. Edgerton VR, Courtine G, Gerasimenko YP, Lavrov I, Ichiyama RM, Fong AJ, et al. Training locomotor networks. Brain Res Rev. 2008;57(1):241–54. 2007/11/21.

    Article  PubMed  Google Scholar 

  11. Jayaraman A, Shah P, Gregory C, Bowden M, Stevens J, Bishop M, et al. Locomotor training and muscle function after incomplete spinal cord injury: case series. JSpinal CordMed. 2008;31(2):185–93.

    Google Scholar 

  12. Harkema SJ, Hillyer J, Schmidt-Read M, Ardolino E, Sisto S, Behrman A. Locomotor training: as a treatment of spinal cord injury and in the progression of neurological rehabilitation. Arch Phys Med Rehabil. 2012;93:1588–97.

    Article  PubMed  Google Scholar 

  13. Lemay JF, Nadeau S. Standing balance assessment in ASIA D paraplegic and tetraplegic participants: concurrent validity of the Berg balance scale. Spinal Cord. 2010;48(3):245–50.

    Article  PubMed  Google Scholar 

  14. Datta S, Lorenz D, Harkema SJ. Dynamic longitudinal evaluation of the utility of the Berg balance scale in individuals with motor incomplete spinal cord injury. Arch Phys Med Rehabil. 2012;93:1565–73.

    Article  PubMed  Google Scholar 

  15. Wirz M, Muller R, Bastiaenen C. Falls in persons with spinal cord injury: validity and reliability of the Berg balance scale. NeurorehabilNeural Repair. 2010;24(1):70–7.

    Article  Google Scholar 

  16. Tse CM, Chisholm AE, Lam T, Eng JJ, Team SR. A systematic review of the effectiveness of task-specific rehabilitation interventions for improving independent sitting and standing function in spinal cord injury. J Spinal Cord Med [Internet]. 20170724th ed. 2018;41(3):254–66. Available from: https://www.ncbi.nlm.nih.gov/pubmed/28738740

  17. Walia S, Kumar P, Kataria C. Interventions to Improve Standing Balance in Individuals With Incomplete Spinal Cord Injury: A Systematic Review and Meta-Analysis. Top Spinal Cord Inj Rehabil [Internet]. 20230522nd ed. 2023;29(2):56–83. Available from: https://www.ncbi.nlm.nih.gov/pubmed/37235196

  18. Jaeger RJ, Yarkony FM, Roth EJ. Rehabilition technology for standing and walking after spinal cord injury. AmJPhysiolMedRehab. 1989;68:128–33.

    CAS  Google Scholar 

  19. Mundra A, Varma Kalidindi K, Chhabra HS, Manghwani J. Spinal cord stimulation for spinal cord injury– Where do we stand? A narrative review. J Clin Orthop Trauma. 2023;43.

  20. Rejc E, Angeli C, Harkema S. Effects of Lumbosacral Spinal Cord Epidural Stimulation for Standing after Chronic Complete Paralysis in Humans. PLoS One. 2015/07/25. 2015;10(7):e0133998.

  21. Rejc E, Angeli CA, Atkinson D, Harkema SJ. Motor recovery after activity-based training with spinal cord epidural stimulation in a chronic motor complete paraplegic. Sci Rep [Internet]. 2017;7(1):13476. Available from: https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41598-017-14003-w

  22. Smith AC, Angeli CA, Ugiliweneza B, Weber KA 2nd, Bert RJ, Negahdar M, et al. Spinal cord imaging markers and recovery of standing with epidural stimulation in individuals with clinically motor complete spinal cord injury. Exp Brain Res. 2022;240(1):279–88. 20211202nd ed.

    Article  PubMed  Google Scholar 

  23. Rejc E, Angeli CA, Bryant N, Harkema SJ. Effects of Stand and Step Training with Epidural Stimulation on Motor Function for Standing in Chronic Complete Paraplegics. J Neurotrauma [Internet]. 2017;34(9):1787–802. Available from: https://www.ncbi.nlm.nih.gov/pubmed/27566051

  24. Angeli C, Rejc E, Boakye M, Herrity A, Mesbah S, Hubscher C et al. Targeted Selection of Stimulation Parameters for Restoration of Motor and Autonomic Function in Individuals With Spinal Cord Injury. Neuromodulation Technol Neural Interface [Internet]. 2023; Available from: https://www.sciencedirect.com/science/article/pii/S1094715923001484

  25. Mesbah S, Gonnelli F, Angeli CA, El-Baz A, Harkema SJ, Rejc E. Neurophysiological markers predicting recovery of standing in humans with chronic motor complete spinal cord injury. Sci Rep. 2019/10/11. 2019;9(1):14474.

  26. Cohen J. Statistical power analysis for the behavioral sciences [Internet]. 2nd ed. Hillsdale, N.J.: L. Erlbaum Associates; 1988. Available from: http://www.gbv.de/dms/bowker/toc/9780805802832.pdf, http://www.catdir.loc.gov/catdir/enhancements/fy0731/88012110-d.html, http://www.zentralblatt-math.org/zmath/en/search/?an=0747.62110

  27. Farivar SS, Liu H, Hays RD. Half standard deviation estimate of the minimally important difference in HRQOL scores? Expert Rev Pharmacoecon Outcomes Res [Internet]. 2004;4(5):515–23. Available from: https://www.ncbi.nlm.nih.gov/pubmed/19807545

  28. Norman GR, Sloan JA, Wyrwich KW. The truly remarkable universality of half a standard deviation: confirmation through another look. Expert Rev Pharmacoecon Outcomes Res [Internet]. 2004;4(5):581–5. Available from: https://www.ncbi.nlm.nih.gov/pubmed/19807551

  29. Haisma JA, van der Woude LH, Stam HJ, Bergen MP, Sluis TA, Bussmann JB. Physical capacity in wheelchair-dependent persons with a spinal cord injury: a critical review of the literature. Spinal Cord. 2006;44(11):642–52.

    Article  CAS  PubMed  Google Scholar 

  30. Martin Ginis KA, van der Ploeg HP, Foster C, Lai B, McBride CB, Ng K et al. Participation of people living with disabilities in physical activity: a global perspective. Lancet [Internet]. 20210721st ed. 2021;398(10298):443–55. Available from: https://www.ncbi.nlm.nih.gov/pubmed/34302764

  31. van den Berg-Emons RJ, Bussmann JB, Stam HJ. Accelerometry-based activity spectrum in persons with chronic physical conditions. Arch Phys Med Rehabil [Internet]. 2010;91(12):1856–61. Available from: https://www.ncbi.nlm.nih.gov/pubmed/21112426

  32. Teasell RW, Arnold JM, Krassioukov A, Delaney GA. Cardiovascular consequences of loss of supraspinal control of the sympathetic nervous system after spinal cord injury. ArchPhysMedRehabil. 2000;81(4):506–16.

    CAS  Google Scholar 

  33. Wecht JM, Bauman WA. Implication of altered autonomic control for orthostatic tolerance in SCI. Aut Neurosci. 2018;209:51–8. 2017/05/14.

    Article  Google Scholar 

  34. Grigorean VT, Sandu AM, Popescu M, Iacobini MA, Stoian R, Neascu C et al. Cardiac dysfunctions following spinal cord injury. J Med Life [Internet]. 2009;2(2):133–45. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20108532

  35. Aslan SC, Legg Ditterline BE, Park MC, Angeli CA, Rejc E, Chen Y et al. Epidural Spinal Cord Stimulation of Lumbosacral Networks Modulates Arterial Blood Pressure in Individuals With Spinal Cord Injury-Induced Cardiovascular Deficits. Front Physiol [Internet]. 2018;9:565. Available from: https://www.ncbi.nlm.nih.gov/pubmed/29867586

  36. Gill M, Linde M, Fautsch K, Hale R, Lopez C, Veith D, et al. Epidural electrical stimulation of the lumbosacral spinal cord improves trunk stability during seated reaching in two humans with severe thoracic spinal cord injury. Front Syst Neurosci. 2020;14:79. 2020/12/18.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Joshi K, Rejc E, Ugiliweneza B, Harkema SJ, Angeli CA. Spinal cord epidural stimulation improves lower spine sitting posture following severe cervical spinal cord injury. Bioengineering. 2023;10(9):1–17.

    Article  Google Scholar 

  38. Bowersock CD, Pisolkar T, Omofuma I, Luna T, Khan M, Santamaria V, et al. Robotic upright stand trainer (RobUST) and postural control in individuals with spinal cord injury. J Spinal Cord Med. 2023;46(6):889–99.

    Article  PubMed  Google Scholar 

  39. Beloozerova IN, Zelenin PV, Popova LB, Orlovsky GN, Grillner S, Deliagina TG. Postural control in the rabbit maintaining balance on the tilting platform. J Neurophysiol [Internet]. 2003;90(6):3783–93. Available from: http://www.ncbi.nlm.nih.gov/pubmed/12930819

  40. Deliagina T, Beloozerova IN, Popova LB, Sirota MG, Swadlow HA, Grant G et al. Role of different sensory inputs for maintenance of body posture in sitting rat and rabbit. Motor Control. 2000/10/06. 2000;4(4):439–52.

  41. Monte A, Benamati A, Pavan A, d’Avella A, Bertucco M. Muscle synergies for multidirectional isometric force generation during maintenance of upright standing posture. Exp Brain Res. 2024;242(8):1881–902.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Gill ML, Grahn PJ, Calvert JS, Linde MB, Lavrov IA, Strommen JA et al. Neuromodulation of lumbosacral spinal networks enables independent stepping after complete paraplegia. Nat Med. 2018.

  43. Darrow DP, Balser DY, Freeman D, Pelrine E, Krassioukov A, Phillips A, et al. Effect of epidural spinal cord stimulation after chronic spinal cord injury on volitional movement and cardiovascular function: study protocol for the phase II open label controlled E-STAND trial. BMJ Open. 2022;12(7):e059126. 20220718th ed.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Grahn PJ, Lavrov IA, Sayenko DG, Van Straaten MG, Gill ML, Strommen JA et al. Enabling Task-Specific Volitional Motor Functions via Spinal Cord Neuromodulation in a Human With Paraplegia. Mayo Clin Proc. 2017/04/08. 2017;92(4):544–54.

  45. Capogrosso M, Wenger N, Raspopovic S, Musienko P, Beauparlant J, Bassi LL, et al. A computational model for epidural electrical stimulation of spinal sensorimotor circuits. JNeurosci. 2013;33(49):19326–40.

    Article  CAS  Google Scholar 

  46. Jilge B, Minassian K, Rattay F, Pinter MM, Gerstenbrand F, Binder H, et al. Initiating extension of the lower limbs in subjects with complete spinal cord injury by epidural lumbar cord stimulation. ExpBrain Res. 2004;154(3):308–26.

    CAS  Google Scholar 

  47. Harkema SJ, Legg Ditterline B, Wang S, Aslan S, Angeli CA, Ovechkin A et al. Epidural Spinal Cord Stimulation Training and Sustained Recovery of Cardiovascular Function in Individuals With Chronic Cervical Spinal Cord Injury. JAMA Neurol [Internet]. 2018;75(12):1569–71. Available from: https://www.ncbi.nlm.nih.gov/pubmed/30242310

Download references

Acknowledgements

We thank the research volunteers for their valuable contribution to this study. We also gratefully acknowledge our research staff for their contribution to the data collection, and our training staff for their support of the research volunteers; Drs Darryl Kaelin, and Sarah Wagers for medical oversight; Yukishia Austin, Lynn Robbins, Kayla Wesley, and Kristen Johnson for medical management.

Funding

This study was funded by the Christopher and Dana Reeve Foundation, Kessler Foundation, Leona M. & Harry B. Helmsley Charitable Trust, University of Louisville Hospital-UofL Health. Medtronic Plc, donated all the devices used in this study. Funders had no role in the study design, data analysis or decision to publish results.

Author information

Authors and Affiliations

  1. Tim and Caroline Reynolds Center for Spinal Stimulation, Kessler Foundation, West Orange, NJ, USA

    Claudia A. Angeli, Enrico Rejc, Gail F. Forrest & Susan J. Harkema

  2. Kentucky Spinal Cord Injury Research Center, University of Louisville, Louisville, KY, USA

    Claudia A. Angeli, Enrico Rejc, Beatrice Ugiliweneza, Maxwell Boakye, Katelyn Brockman, Justin Vogt, Brittany Logsdon & Katie Fields

  3. Department of Bioengineering, University of Louisville, Louisville, KY, USA

    Claudia A. Angeli

  4. Department of Medicine, University of Udine, Udine, Italy

    Enrico Rejc

  5. Department of Anatomical Sciences and Neurobiology, University of Louisville, Louisville, KY, USA

    Beatrice Ugiliweneza

  6. Department of Neurological Surgery, University of Louisville, Louisville, KY, USA

    Maxwell Boakye

Authors
  1. Claudia A. Angeli
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Enrico Rejc
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Beatrice Ugiliweneza
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Maxwell Boakye
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Gail F. Forrest
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Katelyn Brockman
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Justin Vogt
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Brittany Logsdon
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Katie Fields
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Susan J. Harkema
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Conceptualization: CAA, SJH, Methodology: CAA, ER, SJH, Investigation: CAA, ER, MB, GF, KB, JV, BL, KF, SJH, Visualization: CAA, ER, Data curation: CAA, ER, BU, Formal analysis: BU, Funding acquisition: SJH, Supervision: SHJ, CAA, ER, KB, JV, BL, KF, Writing– original draft: CAA, Writing– review & editing: CAA, ER, BU, MB, GF, KB, JV, BL, KF, SJH.

Corresponding author

Correspondence to Claudia A. Angeli.

Ethics declarations

Ethics approval and consent to participate

The Institutional Review Boad of the University of Louisville approved the study protocol (16–0179 and approved on 10/19/2017). All participants provided written informed consent according to the Declaration of Helsinki.

Consent for publication

Research participants involved in the videos and images provided additional consent for publication.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Supplementary Material 3

Supplementary Material 4

Supplementary Material 5

Supplementary Material 6

Supplementary Material 7

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Angeli, C.A., Rejc, E., Ugiliweneza, B. et al. Activity-based recovery training with spinal cord epidural stimulation improves standing performance in cervical spinal cord injury. J NeuroEngineering Rehabil 22, 101 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12984-025-01636-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12984-025-01636-6

Keywords

Journal of NeuroEngineering and Rehabilitation

ISSN: 1743-0003

Contact us