Gait Rehabilitation Using Functional Electrical Stimulation Induces Changes in Ankle Muscle Coordination in Stroke Survivors: A Preliminary Study

Abstract

Background: Previous studies have demonstrated that post-stroke gait rehabilitation combining functional electrical stimulation (FES) applied to the ankle muscles during fast treadmill walking (FastFES) improves gait biomechanics and clinical walking function. However, there is considerable inter-individual variability in response to FastFES. Although FastFES aims to sculpt ankle muscle coordination, whether changes in ankle muscle activity underlie observed gait improvements is unknown. The aim of this study was to investigate three cases illustrating how FastFES modulates ankle muscle recruitment during walking. Methods: We conducted a preliminary case series study on three individuals (53-70 y; 2 M; 35-60 months post-stroke; 19-22 lower extremity Fugl-Meyer) who participated in 18 sessions of FastFES (3 sessions/week; ClinicalTrials.gov: NCT01668602). Clinical walking function (speed, 6-min walk test, and Timed-Up-and-Go test), gait biomechanics (paretic propulsion and ankle angle at initial-contact), and plantarflexor (soleus)/dorsiflexor (tibialis anterior) muscle recruitment were assessed pre- and post-FastFES while walking without stimulation. Results:Two participants (R1, R2) were categorized as responders based on improvements in clinical walking function. Consistent with heterogeneity of clinical and biomechanical changes commonly observed following gait rehabilitation, how muscle activity was altered with FastFES differed between responders. R1 exhibited improved plantarflexor recruitment during stance accompanied by increased paretic propulsion. R2 exhibited improved dorsiflexor recruitment during swing accompanied by improved paretic ankle angle at initial-contact. In contrast, the third participant (NR1), classified as a non-responder, demonstrated increased ankle muscle activity during inappropriate phases of the gait cycle. Across all participants, there was a positive relationship between increased walking speeds after FastFES and reduced SOL/TA muscle coactivation. Conclusion:Our preliminary case series study is the first to demonstrate that improvements in ankle plantarflexor and dorsiflexor muscle recruitment (muscles targeted by FastFES) accompanied improvements in gait biomechanics and walking function following FastFES in individuals post-stroke. Our results also suggest that inducing more appropriate (i.e., reduced) ankle plantar/dorsi-flexor muscle coactivation may be an important neuromuscular mechanism underlying improvements in gait function after FastFES training, suggesting that pre-treatment ankle muscle status could be used for inclusion into FastFES. The findings of this case-series study, albeit preliminary, provide the rationale and foundations for larger-sample studies using similar methodology.

Keywords: biomechanics; electromyography (EMG); functional electrical stimulation (FES); gait rehabilitation; neuromechanics; walking.

Figures

Figure 1

Changes in biomechanics at self-selected walking speed after FastFES. Two biomechanical outcomes were analyzed, consistent with the biomechanical parameters targeted by plantarflexor and dorsiflexor FES, (A) peak paretic propulsion, defined as peak anterior ground reaction force during stance, and (B) the paretic ankle angle at heel-strike, where positive values correspond to dorsiflexion. Participants R1 and NR1 demonstrated significant increase in paretic leg propulsion after FastFES (A). Both responders (Participants R1 and R2) walking with greater dorsiflexion ankle angles at heel-strike after FastFES (B). Note that to prevent the influence of changes in speed on gait biomechanics, the post-training evaluation was conducted while participants walked at their pre-training self-selected speed. (* denotes p ≤ 0.05).

Figure 2

Ankle muscle activity while walking at self-selected speed before and after FastFES. (A) Soleus (SOL) muscle activity across the gait cycle (left, average ± one standard deviation) and integrated SOL activity during stance and swing phases (right, where stance phase was defined as 0–60% of the gait cycle and swing phase as 61–100% of the gait cycle). After FastFES Participant R1 demonstrated increased SOL recruitment during stance, and Participants R1 and NR1 had increased SOL activity during swing. (B) Tibialis anterior (TA) muscle activity across the gait cycle (left, average ± one standard deviation) and integrated TA activity during stance and swing phases (right). After FastFES, Participant R1 decreased TA activity during swing, Participants R2 increased TA activity during swing, and Participant NR1 had increased TA activity during stance. Note that to prevent the influence of changes in speed on muscle activity, post-training trials were selected with speeds matched to the pre-training self-selected walking speed. The average walking speeds across pre-training SSWS trials in participants R1, R2, and NR1 were 0.78 ± 0.07 m/s, 1.60 ± 0.20, and 0.47 ± 0.04, respectively, and for post-training at matched speeds were 0.75 ± 0.03 m/s, 1.70 ± 0.09, and 0.43 ± 0.02, respectively (* denotes p ≤ 0.05). Typical timing is from Perry (26).

Figure 3

Changes in muscle activity modulation across walking speeds after FastFES for (A) Soleus (SOL) during stance and (B) Tibialis Anterier (TA) during swing. Linear regressions between iEMG and walking speed were generated before and after FastFES to examine the degree to which muscle activity was modulated with walking speed. A single regression equation was identified for each muscle in each participant, where the variable speed is walking speed minus the average self-selected walking speed at pre-training (light-colored triangles). Analysis was focused on whether the slopes of the regression for pre-training and post-training were significantly different (H0: βpost−speed = 0). Note that filled markers denote those trials with similar speeds pre- and post-training that were used in Figure 2. Participant R1 increased and decreased modulation of SOL and TA across walking, respectively. Participant R2 increased modulation of TA across walking speeds. Participant NR1 showed no changes in muscle activity modulation after FastFES (SWS, slow walking speed; SSWS, self-selected walking speed; FWS, fast walking speed).

Figure 4

Changes in ankle muscle coactivation after FastFES. (Left) Ankle muscle coactivation between SOL and TA during the gait cycle while walking at slow-walking speeds (SWS), self-selected walking speeds (SSWS), and fast walking speeds (FWS) pre- and post-training. Coactivation was calculated as the percentage of overlapping muscle activity as in Winter 1990. Participants R1 and R2 demonstrated decreases in coactivation after FastFES, whereas Participant NR1 increased coactivation. (Right) Average change in coactivation vs. average change in walking speed at each speed condition. Correlation analysis revealed a trend for increased walking speeds post-training to be associated with reduced SOL/TA muscle coactivation (r = −0.67, p = 0.07). *p < 0.05.