Enhancing Neuroplasticity in the Chronic Phase After Stroke: Effects of a Soft Robotic Exosuit on Training Intensity and Brain-Derived Neurotrophic Factor

Abstract

Objective: High intensity training may enhance neuroplasticity after stroke; however, gait deficits limit the ability to achieve and sustain high walking training intensities. We hypothesize that soft robotic exosuits can facilitate speed-based gait training at higher intensities and longer durations, resulting in a corresponding increase in circulating brain-derived neurotrophic factor (BDNF). Results: Eleven individuals >6-mo post-stroke completed a two-session, pilot randomized crossover trial (NCT05138016). Maximum training speed (Δ: 0.07 ± 0.03 m/s), duration (Δ: 2.07 ± 0.88 min), and intensity (VO₂ peak, Δ: 1.75 ± 0.60 ml-O₂/kg/min) significantly increased (p < 0.05) during exosuit-augmented training compared to no-exosuit training. Post-session increases in BDNF (Δ: 5.96 ± 2.27 ng/ml, p = 0.03) were observed only after exosuit-augmented training. Biomechanical changes were not observed after exosuit-augmented training; however, a deterioration in gait propulsion symmetry (%Δ: -5 ± 2 %) and an increase in nonparetic propulsion (Δ: 0.9 ± 0.3 %bw) were observed (p < 0.05) after no-exosuit training. Conclusion: Soft robotic exosuits facilitate faster, longer duration, and higher intensity walking training associated with enhanced neuroplasticity.

Keywords: Brain-derived neurotrophic factor; neuroplasticity; soft robotic exosuit; stroke; walking.

Figure 1.

Soft Robotic Exosuit, ReWalk ReStore suit components. The ReStore is a textile-based exosuit that provides mechanical assistance to the paretic plantarflexors and dorsiflexors during walking.

Figure 2.

Session overview. Immediately before and after a continuous, speed-graded, maximal effort exercise bout, 10 ml blood samples were collected to enable comparative measurements of baseline and post-exercise serum BDNF levels. Before and after the exercise bout, subjects also completed a 2-minute biomechanical evaluation on an instrumented treadmill (without exosuit assistance) at their comfortable walking speed (CWS). The measured ground reaction force data were used to investigate pre-to-post training changes in interlimb peak propulsion symmetry. Indirect calorimetry and electrocardiogram data were collected during each exercise bout to determine between-condition differences in peak oxygen consumption (VO₂ peak) and monitor subjects, respectively.

Figure 3.

Between-condition comparison of training intensity and duration. Average (a) VO₂ peak and (b) training duration for each training condition. The number of outcome-specific responders (blue) and non-responders (black) is shown to the right of each panel. VO₂ responders were those with a between-condition (exosuit - control) increase in VO₂ peak that was larger than the MDC (2.2 ml-O₂/kg/min) (shown in blue). Training duration responders were those with a between-condition (exosuit – control) increase in duration of any magnitude. Significance (*) at p < 0.05.

Figure 4.

Serum BDNF levels pre-versus-post training. Average pre- and post-training serum BDNF levels in the (a) no-exosuit and (b) exosuit-augmented training conditions. The right of each panel also shows the number of BDNF responders—defined as those with a training-induced change in serum BDNF that exceeded the MDC-UCL₉₅ (9.15 ng/ml) (shown in blue)—and non-responders (shown in black). Significance (*) at p < 0.05.

Figure 5.

Propulsion symmetry pre-versus-post training. Average pre- and post-training (a) interlimb propulsion symmetry and (b) change in individual limb peak propulsion pre-to-post-training in the no-exosuit and exosuit-augmented training conditions. Significance (*) at p < 0.05 Abbreviation: %bw - %bodyweight.