Abstract
Objective:
While cycling and walking are different motor tasks, both require the rapid reciprocal activation and relaxation of lower extremity muscles in a synergistic manner, promoting coordination patterns similar to walking. The aim of this secondary analysis was to examine the effects of forced-rate cycling on gait biomechanics after stroke.
Design:
A randomized controlled trial was conducted investigating the neuroplastic effects of forced-rate cycling on upper extremity motor recovery in individuals (n = 60) after stroke. In this secondary analysis, a subset of participants (N = 45) completed biomechanical gait assessment at baseline and postintervention.
Results:
Gait velocity increased in the cycling group by 0.08 m/sec and in the control group by 0.02 m/sec, though a group difference was not observed (P = 0.195). Walking capacity (6-min walk test) significantly improved by 53.3 m for the cycling group versus 15.6 m for the control group (P = 0.0036). Improvements in spatiotemporal characteristics in the cycling group included increased step length, cadence, and time spent in single-limb support, in addition to paretic limb kinetics.
Conclusions:
Improvements in gait velocity after forced-rate cycling were accompanied by normalization of gait biomechanics rather than exaggerated compensatory strategies, supporting the use of aerobic cycling as a viable and safe intervention to improve locomotor control after stroke.
Keywords: Walking, Locomotor Training, Hemiparesis, Gait Disorders, Recovery of Function
The recovery of walking has been reported as particularly meaningful by individuals with stroke, as walking-related disability negatively impacts participation in daily activities, community reintegration, and quality of life.1 A recently published clinical practice guideline reported that motor learning-based approaches focused primarily on task-specific gait training with and without the use of technology were most effective in restoring walking ability after stroke.1 The most effective interventions reported were moderate- to high-intensity and virtual reality-based walking programs. Interestingly, despite numerous well-designed clinical trials investigating the use of body weight-supported treadmill training and robotic-assisted gait training, neither were recommended, citing a preponderance of risks, harm, and costs.1,2 Additionally, the guideline cited insufficient evidence for cycling interventions to improve walking capacity, although high-intensity cycling interventions were found to be more efficacious compared with low intensity cycling.1
In contrast to findings of the clinical guideline, while studying the effects of aerobic cycling on upper extremity motor recovery, we have reported improvements in gait velocity and walking capacity after cycling interventions.3–6 Additionally, we have demonstrated that improvements in gait velocity were accompanied by improved biomechanics in a small cohort of individuals participating in an 8-wk forced-rate aerobic cycling intervention.5 Notably, increases in gait velocity were accompanied by improved locomotor control, evidenced by the use of gait strategies that were closer to able-bodied, neurologically healthy adults. We theorized that high intensity cycling may train muscle groups to work synergistically to ensure smooth intralimb and interlimb reciprocal activation, similar to activation patterns used to coordinate joint angle accelerations and decelerations during phases of the gait cycle.7–11
The mode of aerobic cycling we investigated was forced-rate exercise (FE). During FE, the voluntary efforts of individuals are supplemented with a motor, allowing them to achieve and sustain a greater exercise rate and intensity than what they can achieve without assistance.3,12–14 Importantly, the participants’ efforts are supplemented, but not replaced, as heart rate (HR) and pedaling torques are monitored continuously to ensure that participants are contributing to the exercise and elicit an aerobic response. The aim of the larger trial was to examine the neuroplastic effects of FE on upper limb recovery. However, as we had previously observed improvements in walking capacity after FE, an exploratory aim was to determine the effects of FE on gait biomechanics by comparing changes in spatiotemporal, kinematic, and kinetic components of gait in individuals with chronic stroke completing an 8-wk FE intervention to a non-cycling control group. We also sought to examine exercise-related variables that were predictive of increased gait velocity among FE participants. We hypothesized that those participating in FE would demonstrate increases in gait velocity accompanied by normalization of spatiotemporal, kinematic, and kinetic parameters of gait, while no significant improvements in gait were expected for the control group.
METHODS
This study was part of a larger randomized clinical trial aimed at determining effects of an 8-wk (3×/wk) FE intervention combined with upper extremity repetitive task practice (UE RTP) on UE motor recovery compared with time-matched UE RTP only (clinicaltrials.gov registration number NCT03819764).4,15 Given observations from our pilot studies that FE elicited improvements in walking capacity,3 biomechanical gait data were collected using three-dimensional (3D) motion capture as an exploratory aim. The study was approved by the Cleveland Clinic Institutional Review Board and all participants provided informed consent.
Participants
Individuals ≥6 mos after a single, unilateral ischemic or hemorrhagic stroke with residual UE hemiparesis were recruited for participation, as the primary aim of the study was to determine the effects of FE on UE recovery. Additional inclusion/exclusion criteria have been previously published.4,15 All enrolled participants underwent a metabolic exercise stress test on an electronically controlled cycle ergometer to ensure safe cardiopulmonary response to maximal exertion15 and to determine target HR and cadence parameters for those randomized to FE. Following baseline exercise stress and clinical testing, participants underwent concealed randomization stratified by baseline upper extremity Fugl-Meyer level (≤33 or >33) using a nonreplenished sealed opaque envelope pull method.
Forced Exercise and Upper Extremity Repetitive Task Practice
The FE protocol was based on methodology from our previous studies3,12–14 and involved supervised and monitored exercise on a custom-engineered stationary semi-recumbent cycle ergometer equipped with a motor that augmented pedaling rate by 30% greater than the participant’s voluntary rate. Target HR zone was determined using the Karvonen formula at the 60%–80% range, based on values obtained during the baseline stress test. Sessions included a 5-min warm-up, 35-min main exercise set, and 5-min cool down. Participants were instructed to exercise within their target HR zone during the 35-min main exercise set. Heart rate was measured continuously using a Wahoo chest strap (Wahoo, Atlanta, GA). Cycling shoes with cleats were used to ensure a secure interaction between the individual’s feet and pedals, and training sessions were conducted without ankle foot orthoses.
Upper Extremity Repetitive Task Practice
Repetitive task practice is considered the current standard of care for UE stroke rehabilitation, with Class IA evidence supporting its use.16 Tasks performed with the more impaired UE that require a combination of reaching, grasping, manipulating and/or moving, and releasing an object were identical to the approach used in our preliminary studies.4 The FE + RTP group completed a 45-min session of FE followed by 45 mins of UE RTP, while the control group completed 90 mins of UE RTP.
Gait Analysis
The Computer Assisted Rehabilitation ENvironment (CAREN) (Motekforce Link, Amsterdam, the Netherlands) system was used to collect biomechanical gait data at baseline and following the 8-wk intervention. The CAREN system engineer, blinded to group allocation, was responsible for all aspects of gait data collection. The CAREN system consists of a 10-camera 3D motion capture system (Vicon Inc., Oxford, UK), D-Flow control software (Motekforce Link), 180° curved projection screen, and a six degree of freedom motion platform (Moog Inc., Elma, NY) with a force-instrumented split-belt treadmill (Bertec Corp., Columbus, OH). Twenty-six retroreflective markers were placed on anatomic landmarks on the lower extremities and trunk as defined by the Human Body Model 2.17 Analog and marker position data were filtered using a second-order low-pass Butterworth filter with a 6-Hz cutoff frequency.
Gait Outcomes
The primary outcome was comfortable gait velocity from baseline to end of treatment (EOT). All walking trials were completed at a comfortable fixed speed on the treadmill. To determine comfortable gait velocity, treadmill speed was gradually increased during a practice trial until the participant stated that he/she was at a comfortable pace, then slightly increased to verify that the comfortable pace was not underestimated; at which point, the treadmill speed was fixed for the duration of data collection. Once a comfortable speed was established, two 2-min trials were completed. A 2- to 3-min rest was provided between trials. The initial walk to determine comfortable velocity and the first 2-min trial were used to acclimatize participants to the gait analysis setup. Gait data from the second trial were used for analysis.
The 6-min walk test (6MWT) was used to assess walking capacity during overground ambulation. Participants were asked to self-select a brisk but safe walking speed to cover their greatest distance along a 200-foot (61 meter) oval path in 6 mins with a trained assessor guarding against loss of balance while ensuring to not influence self-selected walking pace. The assessor for all clinical tests and for the exercise stress test were blinded to group allocation.
Additional secondary outcomes included spatiotemporal, kinematic and kinetic gait parameters. The following spatiotemporal variables were computed: cadence, paretic and nonparetic limb step length, and paretic and nonparetic stance and swing percentage. Sagittal plane kinematics (i.e., joint angles) were of greatest interest; therefore, hip flexion/extension, knee flexion/extension, and ankle plantarflexion/dorsiflexion were computed for paretic and nonparetic limbs. The following kinetic variables were computed: peak vertical ground reaction force (GRF), peak anterior-posterior breaking forces (peak APbraking), peak anterior-posterior propulsion forces (peak APprop), and peak lateral GRF. Hip extension, knee extension, and ankle plantarflexion total positive power (i.e., work) were computed by calculating the positive area under the curve. Peak hip extension moment, knee extension moment and ankle plantarflexion moment during stance were obtained.
Statistical Analysis
Descriptive statistics were computed for demographic characteristics of both groups and exercise variables for the FE + RTP group. Groups were compared on demographic characteristics using analysis of variance, Mann-Whitney U, or χ2 as appropriate. Normality of data was determined using visual inspection of histograms and normal Q-Q plots, along with the Shapiro-Wilk test. The primary outcome, comfortable gait velocity from baseline to EOT, was analyzed using an analysis of covariance, with baseline gait velocity as the covariate and an alpha of 0.05 to determine significance. The 6MWT was also analyzed using an analysis of covariance with baseline 6MWT distance as the covariate. The remaining spatiotemporal, kinematic, and kinetic (GRF, power, and moment) variables were compared using two-way multivariate analysis of variance (MANOVA) linear models. Paretic and nonparetic limb variables were analyzed in separate models. Post-hoc univariate comparisons were conducted using separate t tests. Finally, a multivariate stepwise backwards linear regression was constructed for the FE + RTP group only to determine predictors of greatest improvement in gait velocity with percent change in gait velocity as the dependent variable as exercise variables as independent variables.
RESULTS
Of the 60 participants, 45 completed biomechanical gait analysis at baseline and postintervention (FE + RTP, n = 25; control, n = 20). Participants were 60.4 ± 10.8 yrs of age and a median of 20 (range 6–241) mos after stroke. The consolidated standards of reporting trials diagram is shown in Figure 1. Participant demographics and baseline characteristics summarized by group, and exercise variables for the FE + RTP group are provided in Table 1.
FIGURE 1.
Consolidated standards of reporting trials flow diagram. AE, adverse event; EOT, end of treatment.
TABLE 1.
Participant demographics and exercise characteristics
| Overall (N = 45) | FE + RTP (n = 25) | Control (n = 20) | P | |
|---|---|---|---|---|
| Age (yrs) | 60.4 ± 10.8 | 61.2 ± 12.1 | 59.6 ± 9.28 | 0.615 |
| Female sex (vs. male), n (%) | 20 (44%) | 10 (40%) | 10 (50%) | 0.502 |
| Dominant side affected, n (%) | 22 (49%) | 14 (56%) | 8 (40%) | 0.286 |
| Time since stroke (mos) | 20 (12, 60) | 18 (11, 59) | 26 (13.5, 66.2) | 0.615 |
| Modified Rankin Score, n (%) | ||||
| 1 | 2 (4%) | 1 (4%) | 1 (5%) | 0.840 |
| 2 | 27 (60%) | 15 (60%) | 12 (60%) | |
| 3 | 13 (29%) | 7 (28%) | 6 (30%) | |
| 4 | 1 (2%) | 1 (4%) | 0 (0%) | |
| Stroke type, n (%) | ||||
| Ischemic | 36 (80%) | 21 (84%) | 15 (75%) | 0.453 |
| Hemorrhagic | 9 (20%) | 4 (16%) | 5 (25%) | |
| Height, cm | 170.4 ± 10.9 | 170.0 ± 10.4 | 170.0 ± 11.8 | 0.968 |
| Mass, kg | 84.1 ± 19.6 | 82.6 ± 8.0 | 85.6 ± 21.3 | 0.561 |
| Exercise characteristics | ||||
| Cadence (RPM) | 74.4 ± 7.46 | |||
| Percentage of HRR | 62.2 (53.5, 64.7) | |||
| Power (W) | 48.5 (7.01, 75.2) |
Summary statistics presented as mean ± standard deviation for normally distributed variables, median (Q1 to Q3) for nonnormally distributed variables, or n (%) for categorical data.
HRR, heart rate reserve.
FE Elicits Improvements in Spatiotemporal Parameters of Gait and Walking Capacity
The primary outcome, comfortable gait velocity, for both groups is shown in Table 2 and for all participants in Figure 2A. Those in the FE + RTP group improved from 0.57 to 0.65 m/sec while the control group improved from 0.65 to 0.67 m/sec. The difference in gait velocity improvement was not significant between groups (P = 0.195); however, post-hoc analysis demonstrated the FE + RTP group improved significantly from baseline to EOT (P = 0.0013), whereas improvement for the control group was not significant (P = 0.678). The 6MWT, a secondary outcome, improved by 53.3 m in FE + RTP group versus 15.6 m in the control group (P = 0.0036). In post-hoc analysis, both the FE + RTP and control groups demonstrated significantly increased 6MWT performance (P < 0.001 for FE + RTP, P = 0.046 for control). A bar plot, Figure 2B, depicts the change in 6MWT for all participants.
TABLE 2.
Spatiotemporal parameters of gait
| FE + RTP (n = 25) | Control Group (n = 20) | ||||
|---|---|---|---|---|---|
| Variable | Baseline | EOT | Baseline | EOT | P |
| Primary outcome | |||||
| Gait velocity (m/sec) | 0.57 ± 0.28 | 0.65 ± 0.28 | 0.65 ± 0.32 | 0.67 ± 0.26 | 0.195 |
| Secondary outcome | |||||
| 6MWT (m) | 261.1 ± 124.4 | 314.4 ± 142.9 | 339.0 ± 144.6 | 354.6 ± 142.0 | 0.0036 |
| Spatiotemporal variables: overall MANOVA | 0.211 | ||||
| Cadence (steps/min) | 82.1 ± 17.2 | 86.1 ± 17.3 | 83.8 ± 17.7 | 85.8 ± 15.4 | 0.323 |
| Paretic side | |||||
| Normalized step length (cm) | 40.4 ± 14.1 | 45.0 ± 12.6 | 45.6 ± 12.7 | 46.9 ± 11.5 | 0.025 |
| Single limb % | 26.8 ± 3.88 | 27.9 ± 3.40 | 28.6 ± 2.76 | 28.9 ± 2.70 | 0.058 |
| Nonparetic side | |||||
| Normalized step length (cm) | 40.1 ± 15.5 | 45.0 ± 15.3 | 44.2 ± 14.9 | 46.1 ± 11.5 | 0.108 |
| Single limb % | 29.1 ± 4.79 | 29.8 ± 4.73 | 30.1 ± 2.64 | 30.5 ± 2.30 | 0.058 |
Summary statistics presented as mean ± standard deviation for normally distributed variables.
Italicized values denote improvement in variable from baseline to EOT.
FIGURE 2.
Bar plots depicting change in gait velocity (2A) and change in walking capacity measured by 6-min walk test distance (2B) for FE + RTP participants (black bars) and RTP only (control) participants (light gray bars).
Spatiotemporal parameters of gait are shown in Table 2. When analyzing change in spatiotemporal parameters between groups, using Pillai’s trace, the main effect was not significant, V = 0.16, F(5, 39) = 1.50, P = 0.211. Univariate analyses revealed that those in the FE + RTP group demonstrated improvements in gait cadence as well as paretic and nonparetic limb step length. Improvement in paretic step length was significantly greater in the FE + RTP group compared with the control group (P = 0.025).
Effects of FE on Kinematic Parameters of Gait
Average sagittal plane range of motion for the hip, knee, and ankle in both paretic and nonparetic limbs is shown in Table 3. A modest increase in range of motion in all joints was observed postintervention for the FE + RTP group, while the control group demonstrated no change or slight decrease in joint kinematics. The MANOVA did not reveal a significant main effect in the paretic (V = 0.083, F(3, 41) = 1.349, P = 0.270) or nonparetic limbs (V = 0.083, F(3, 41) = 1.243, P = 0.306).
TABLE 3.
Kinematic parameters of gait
| FE + RTP (n = 25) | Control (n = 20) | ||||
|---|---|---|---|---|---|
| Baseline | EOT | Baseline | EOT | P | |
| Kinematic variables: overall MANOVA | |||||
| Paretic side | 0.270 | ||||
| Hip flex/ext, degrees of ROM | 33.2 ± 9.39 | 35.6 ± 10.0 | 37.4 ± 10.4 | 39.5 ± 7.64 | 0.471 |
| Knee flex/ext, degrees of ROM | 43.0 ± 16.9 | 44.0 ± 17.5 | 54.1 ± 14.4 | 53.4 ± 12.5 | 0.252 |
| Ankle dorsi-plantarflexion, degrees of ROM | 17.8 ± 5.38 | 19.1 ± 5.46 | 19.5 ± 6.45 | 19.1 ± 6.10 | 0.117 |
| Nonparetic side | 0.306 | ||||
| Hip flex/ext, degrees of ROM | 36.9 ± 9.35 | 39.9 ± 9.37 | 41.3 ± 6.14 | 43.0 ± 5.89 | 0.564 |
| Knee flex/ext, degrees of ROM | 53.6 ± 11.8 | 54.8 ± 10.9 | 60.5 ± 9.12 | 62.4 ± 6.41 | 0.612 |
| Ankle dorsi-plantarflexion, degrees of ROM | 23.3 ± 8.74 | 24.8 ± 8.48 | 23.7 ± 7.28 | 23.2 ± 6.89 | 0.088 |
Summary statistics presented as mean ± standard deviation for normally distributed variables median (Q1 to Q3) for nonnormally distributed variables, or n (%) for categorical data.
Italicized values denote improvement in variable from baseline to EOT.
ROM, range of motion.
Effects of FE on Kinetic Gait Characteristics
Data for GRF, joint power, and joint moments for both groups are shown in Table 4. Overall, the FE + RTP group demonstrated improvements driven largely by changes in paretic limb kinetics. None of the analyses involving the nonparetic limb were significantly different between groups.
TABLE 4.
Kinetic parameters of gait
| FE + RTP (n = 25) | Control (n = 20) | ||||
|---|---|---|---|---|---|
| Baseline | EOT | Baseline | EOT | P | |
| Paretic limb | |||||
| Ground reaction forces (N) | 0.079 | ||||
| Peak vGRF | 727 ± 174 | 736 ± 187 | 823 ± 230 | 822 ± 267 | 0.568 |
| Peak APpropulsion | −50.2 ± −31.4 | −58.9 ± −34.8 | −68.3 ± −59.8 | −68.1 ± −60.4 | 0.124 |
| Peak APbraking | 61.9 ± 37.8 | 69.1 ± 39.0 | 76.3 ± 46.8 | 83.4 ± 53.9 | 0.992 |
| Peak lateral GRF | 8.95 ± 10.8 | 11.3 ± 12.0 | 14.3 ± 9.73 | 12.0 ± 7.24 | 0.040 |
| Total power generation in stance (J) | 0.043 | ||||
| Hip | 13.0 ± 11.3 | 15.6 ± 11.9 | 17.4 ± 11.5 | 16.8 ± 8.53 | 0.181 |
| Knee | 12.1 ± 10.1 | 12.4 ± 9.17 | 14.6 ± 11.3 | 15.3 ± 12.7 | 0.890 |
| Ankle | 9.66 ± 7.59 | 11.5 ± 7.98 | 14.5 ± 12.2 | 14.4 ± 11.5 | 0.112 |
| Joint moment (Nm/kg) | 0.111 | ||||
| Hip extension | 0.19 ± 0.15 | 0.25 ± 0.17 | 0.26 ± 0.18 | 0.22 ± 0.11 | 0.031 |
| Knee extension | 0.76 ± 0.49 | 0.60 ± 0.34 | 0.73 ± 0.46 | 0.69 ± 0.40 | 0.288 |
| Ankle plantarflexion | 0.79 ± 0.46 | 1.00 ± 0.46 | 0.88 ± 0.44 | 0.96 ± 0.39 | 0.106 |
| Nonparetic limb | |||||
| Ground reaction forces (N) | 0.532 | ||||
| Peak vGRF | 781 ± 176 | 784 ± 175 | 806 ± 214 | 812 ± 230 | 0.840 |
| Peak APpropulsion | −72.6 ± −36.4 | −83.8 ± −36.3 | −95.0 ± −49.6 | −98.7 ± −43.3 | 0.146 |
| Peak APbraking | 43.3 ± 32.7 | 53.2 ± 29.4 | 52.1 ± 49.7 | 55.2 ± 54.4 | 0.308 |
| Peak lateral GRF | 15.4 ± 13.5 | 19.6 ± 14.2 | 17.7 ± 16.9 | 18.0 ± 18.0 | 0.559 |
| Total power generation in stance (J) | 0.835 | ||||
| Hip | 23.0 ± 21.8 | 25.4 ± 21.6 | 15.3 ± 8.37 | 18.9 ± 10.9 | 0.774 |
| Knee | 15.1 ± 14.3 | 15.9 ± 13.9 | 13.4 ± 8.08 | 16.2 ± 19.3 | 0.456 |
| Ankle | 17.7 ± 12.0 | 19.2 ± 11.7 | 17.0 ± 11.0 | 17.8 ± 9.60 | 0.662 |
| Joint moment (Nm/kg) | 0.447 | ||||
| Hip extension | 0.29 ± 0.22 | 0.37 ± 0.21 | 0.33 ± 0.18 | 0.35 ± 0.17 | 0.107 |
| Knee extension | 0.74 ± 0.46 | 0.73 ± 0.41 | 0.78 ± 0.48 | 0.80 ± 0.42 | 0.856 |
| Ankle plantarflexion | 1.03 ± 0.39 | 1.06 ± 0.40 | 1.03 ± 0.51 | 1.04 ± 0.50 | 0.795 |
Summary statistics presented as mean ± standard deviation for normally distributed variables median (Q1 to Q3) for nonnormally distributed variables, or n (%) for categorical data.
Italicized values denote improvement in variable from baseline to EOT.
vGRF, vertical ground reaction force
Among GRF variables, modest improvements were observed in the paretic limb for FE + RTP group. While the MANOVA did not reveal a significant overall difference between groups, post-hoc univariate analyses revealed significant difference in peak lateral GRF, favoring the FE + RTP group.
Increased total paretic limb hip, knee, and ankle power were observed from baseline to EOT for the FE + RTP group. Using Pillai’s trace, the MANOVA revealed a significant main effect for joint power from baseline to EOT (V = 0.200, F(3, 36) = 3.01, P = 0.043). However, post-hoc univariate models were not significant in total hip, knee, or ankle power generation.
Changes in hip, knee, and ankle joint moments from baseline to EOT for both groups are in Table 4. Using Pillai’s trace, the MANOVA did not revealed a significant group by time effect for change in paretic limb joint moment (V = 0.152, F(3, 36) = 2.15, P = 0.111). Post-hoc univariate linear models revealed that peak hip extension moment (P = 0.031) had a significant group by time interaction effect.
Exercise-Related Predictors of Greatest Improvement in Gait Velocity
When analyzing which exercise variables (cycling cadence, aerobic intensity, or exercise power) were predictive of greatest improvements in gait velocity, the regression model did not identify a significant relationship between exercise variables and improvements in gait velocity (F6,18 = 0.499, P = 0.801).
DISCUSSION
Our results indicate that the FE cycling intervention elicited increased gait velocity in individuals with chronic stroke by 0.08 m/s, a 14% improvement from baseline. While a statistically significant effect of group was not present, FE elicited significant and potentially clinically relevant improvements, as 20 of the 25 FE + RTP participants demonstrated increased gait velocity with eight exceeding the minimal clinically important difference value of 0.1 m/sec. Notably, these improvements occurred in the absence of task-specific gait training and were accompanied by improvements in spatiotemporal and kinetic characteristics of gait, indicating normalization of gait biomechanics and improvements in locomotor control. Additionally, while biomechanical metrics were obtained during self-selected comfortable velocity on a treadmill, significant improvements in overground walking capacity were observed, with those participating in FE + RTP improving by an average of 53.3 m on the 6MWT, exceeding the minimal clinically important difference value of 34.3 m.18 Collectively, these findings provide rationale to incorporate aerobic cycling interventions as an option to improve locomotor function in persons with chronic stroke. Furthermore, our findings are in contrast to the clinical practice guideline that found weak evidence for cycling to improve locomotor function in individuals with chronic stroke.1
Improvements in Spatiotemporal Characteristics of Gait After FE + RTP
Among spatiotemporal gait characteristics, while only increased paretic limb step length was significantly different between groups, the FE + RTP group demonstrated increased step length bilaterally by nearly 5 cm and increased time spent in single limb support bilaterally. These increases were primarily driven by decreased time spent in stance, as swing times bilaterally did not change appreciably. While it is not uncommon for spatiotemporal variables to increase in proportion to faster gait velocities,19 improvements were seen in both paretic and nonparetic limbs, indicating that individuals participating in FE + RTP demonstrated normalization of gait characteristics instead of exaggerating compensatory strategies or relying to a greater extent on the nonparetic limb to walk faster.19
Kinematic Gait Changes After FE + RTP
Modest improvements in hip, knee, and ankle sagittal plane range of motion for both the paretic and nonparetic limbs were observed in the FE + RTP group. However, the same improvements were not observed in the control group. Investigating kinematics associated with increased gait velocity is important to understand mechanisms undertaken by participants to walk faster. For example, in a study investigating kinetic and kinematic characteristics of gait and their relationship to walking speed in persons with stroke, Eng and Kim found that individuals who had the fastest self-selected walking speed did not exhibit profiles resembling neurologically healthy adults.20 In fact, compensatory strategies at the hip (i.e., use of abduction and circumduction rather than flexion to advance the limb) were more common among those who walked the fastest.20 However, normalizing gait mechanics is important in stroke rehabilitation, in part to prevent secondary impairments or injuries that may derive from poor biomechanics.21 Swing phase compensatory strategies such as hip abduction, circumduction, or hip hiking may cause back or hip pain due to overuse of the trunk and lower limb muscles, soft tissue strain, and biomechanical malalignment.22 Similarly, abnormal stance phase mechanics such as genu recurvatum and decreased knee flexion with loading response can place excessive strain on the spine, hip, knee, ankle, or connective tissue, which over time can cause pain and musculoskeletal pathology.22
Changes in Gait Kinetics After FE + RTP
The analysis of gait kinetics provides a window into the etiology of abnormal movement patterns and underlying muscle and joint malfunction.23 Improvements in paretic limb peak lateral GRF for those participating in FE + RTP are indicative of increased lateral weight shift onto the hemiparetic limb, increased peak loading with weight acceptance and normalization of anterior-posterior breaking forces with loading response and propulsion at terminal stance.24 Hip extension and ankle plantarflexion during stance are the main sources of propulsion during gait.25,26 Power and joint moment improved significantly for the FE + RTP group compared with the control group, with greatest improvements observed with total hip extension power, ankle plantarflexion power, and peak hip extension moment during stance. It is important to note that FE was administered without participants wearing ankle-foot orthoses. This allowed for training without restricting ankle movement and may have contributed to increased power of ankle plantar flexors which translated to improved propulsion during gait. Collectively, improvements in gait kinetics after FE + RTP are indicative of improved alignment, improved limb loading, and improved force generation, all of which likely contributed to increased gait velocity.23,27
Cycling May Induce Transfer of Training to Improve Gait
After stroke, individuals present with asymmetric diminished muscle power and abnormal timing and coordination of muscle agonists and antagonists, which disrupt the modulation of phasic muscle activity, resulting in abnormal co-contractions and inefficient movement patterns during gait.19,27 In individuals with poststroke hemiparesis, asymmetrical power generation between the paretic and nonparetic limbs and excessive negative work during flexion phases are distinct characteristics found in both walking and cycling.28 Cycling and walking both involve cyclical movements, alternating flexion and extension at an approximate frequency of 1 Hz, with most of the propulsive power generated through extension.7 Raasch and Zajac described the synergistic action of muscle groups during cycling, finding that the hip extensors, knee extensors, and ankle plantar flexors work synergistically during downstroke, similar to patterns of activation measured during walking.7,28 While the timing and coordination of muscle groups are not identical in cycling and walking, it is plausible that the improvements in power generation were facilitated through a training effect from the FE intervention.
In addition to the similarities with force generation, prolonged excitation of the quadriceps has been found during the up-stroke phase of cycling, resulting in excessive negative work.9,10 This alteration in normal activation patterns during cycling is not dissimilar to prolonged activation and abnormal coactivation of the extensors that is observed during terminal stance through initial swing, often resulting in stiff-legged gait and compensatory strategies for limb advancement.21,22,26 The FE approach involved a controlled pedaling rate and consistent pattern of exercise to train the rapid reciprocal activation and relaxation of lower extremity
Numerous advantages exist to FE training for individuals after stroke including safety, the ability to complete thousands of repetitions in a single session, replicating high cadence associated with normal human gait, and the potential global benefits to aerobic exercise training.29 Seated cycling, particularly on a semirecumbent stationary ergometer, requires less postural control than walking, providing a safe modality to train without individuals having to focus on balance. Additionally, cycling is possible without the use of ankle foot orthoses, which from our results, may have resulted in increased plantarflexion propulsive power translated during gait. Thus, our findings that improvements in gait velocity after FE were accompanied by improvements in gait biomechanics are clinically relevant as these data support the use of FE as a viable and safe aerobic intervention for individuals with chronic stroke to improve locomotor control. In addition to the effects on walking, primary outcomes from this trial revealed that FE + RTP produced similar improvements in UE motor recovery as an RTP intervention that was twice as long.4 Similar motor response with half the RTP dose suggests that adverse event may be priming the central nervous system to optimize neuroplasticity in persons with stroke and should be considered as an adjunct to traditional motor task practice.
Study Limitations
There are limitations to this study; namely, given that the primary aim of the clinical trial was to improve UE function, criteria for participation were related to UE impairment and did not consider the degree to which gait was impaired. Our primary outcome, comfortable gait velocity, was obtained during treadmill walking, which may not be indicative of participants’ gait pattern overground, though similarities between the two approaches have been reported.30 To account for this, an acclimatization period was used and data only after this period were used for analysis. Additionally, while we hypothesize how muscle activation patterns may have been altered as a result of FE training, electromyographic measures were not included to confirm our hypothesis. Lastly, participants were permitted to use ankle foot orthoses as prescribed for community ambulation during gait testing in addition to upper extremity support of one or both handrails. To mitigate this study limitation, the use of orthoses and/or handrails was kept consistent within participants at both testing time points. Therefore, these results may still be used as a basis of comparison between the preintervention and postintervention time points.
Conclusions and Future Directions
Improvements in gait velocity after forced-rate cycling were accompanied by normalization of gait biomechanics rather than exaggerated compensatory strategies, supporting the use of aerobic cycling as a viable and safe intervention to improve locomotor control after stroke. A clinical trial is currently underway to determine the effects of FE in facilitating UE and LE motor recovery postsubacute stroke, to elucidate neural and biochemical substrates of FE-induced motor recovery, and to evaluate cost-effectiveness of a FE-centered intervention.
What Is Known
Current guidelines recommend task-specific gait training for recovery of walking after stroke. The guideline cited weak and insufficient evidence for cycling interventions to improve walking capacity, although high-intensity cycling interventions were found to be more efficacious compared with low intensity cycling.
What Is New
Forced-rate aerobic cycling elicited significant increases in walking capacity and gait velocity accompanied by improvements in spatiotemporal and kinetic characteristics of gait, indicating normalization of gait biomechanics and improvements in locomotor control. Collectively, these findings provide rationale to incorporate aerobic cycling interventions as an option to improve locomotor function in persons with chronic stroke.
ACKNOWLEDGMENTS
The authors thank Ann Bischof-Bockbrader and Anson Rosenfeldt for assisting with the interventions for this trial.
Funding:
The trial was funded by the National Institutes of Health (K01HD092556) to SML.
Footnotes
Competing Interests: Jay Alberts declares a conflict of interest related to intellectual property for the motorized cycle used in the trial. None of the remaining authors report a conflict of interest.
Clinical Trial Registration: clinicaltrials.gov number NCT03819764
Data Sharing:
The full deidentified data set will be available upon request.
REFERENCES
- 1.Hornby TG, Reisman DS, Ward IG, et al. : Clinical practice guideline to improve locomotor function following chronic stroke, incomplete spinal cord injury, and brain injury. J Neurol Phys Ther 2020;44:49–100 [DOI] [PubMed] [Google Scholar]
- 2.Hsu CY, Cheng YH, Lai CH, et al. : Clinical non-superiority of technology-assisted gait training with body weight support in patients with subacute stroke: a meta-analysis. Ann Phys Rehabil Med 2020;63:535–42 [DOI] [PubMed] [Google Scholar]
- 3.Linder SM, Davidson S, Rosenfeldt A, et al. : Forced and voluntary aerobic cycling interventions improve walking capacity in individuals with chronic stroke. Arch Phys Med Rehabil 2021;102:1–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Linder SM, Bischof-Bockbrader A, Davidson S, et al. : The utilization of forced-rate cycling to facilitate motor recovery following stroke: a randomized clinical trial. Neurorehabil Neural Repair 2024;38:291–302 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Linder SM, Learman K, Miller Koop M, et al. : Increased comfortable gait speed is associated with improved gait biomechanics in persons with chronic stroke completing an 8-week forced-rate aerobic cycling intervention: a preliminary study. Am J Phys Med Rehabil 2023;102:619–24 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Khan M, Maag LM, Harnegie MP, et al. : The effects of cycling on walking outcomes in adults with stroke: a systematic review. Top Stroke Rehabil 2024;31:259–71 [DOI] [PubMed] [Google Scholar]
- 7.Raasch CC, Zajac FE: Locomotor strategy for pedaling: muscle groups and biomechanical functions. J Neurophysiol 1999;82:515–25 [DOI] [PubMed] [Google Scholar]
- 8.Raasch CC, Zajac FE, Ma B, et al. : Muscle coordination of maximum-speed pedaling. J Biomech 1997;30:595–602 [DOI] [PubMed] [Google Scholar]
- 9.Kautz SA, Brown DA: Relationships between timing of muscle excitation and impaired motor performance during cyclical lower extremity movement in post-stroke hemiplegia. Brain 1998;121(Pt 3):515–26 [DOI] [PubMed] [Google Scholar]
- 10.Linder SM, Rosenfeldt AB, Bazyk AS, et al. : Improved lower extremity pedaling mechanics in individuals with stroke under maximal workloads. Top Stroke Rehabil 2018;25:248–55 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chen HY, Chen SC, Chen JJ, et al. : Kinesiological and kinematical analysis for stroke subjects with asymmetrical cycling movement patterns. J Electromyogr Kinesiol 2005;15:587–95 [DOI] [PubMed] [Google Scholar]
- 12.Linder SM, Rosenfeldt AB, Davidson S, et al. : Forced, not voluntary, aerobic exercise enhances motor recovery in persons with chronic stroke. Neurorehabil Neural Repair 2019; 33:681–90 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Linder SM, Rosenfeldt AB, Dey T, et al. : Forced aerobic exercise preceding task practice improves motor recovery poststroke. Am J Occup Ther 2017;71:7102290020p1–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Linder SM, Davidson S, Rosenfeldt A, et al. : Predictors of improved aerobic capacity in individuals with chronic stroke participating in cycling interventions. Arch Phys Med Rehabil 2020;101:717–21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Linder SM, Lee J, Bethoux F, et al. : An 8-week forced-rate aerobic cycling program improves cardiorespiratory fitness in persons with chronic stroke: a randomized controlled trial. Arch Phys Med Rehabil 2024;105:835–42 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Winstein CJ, Stein J, Arena R, et al. : Guidelines for adult stroke rehabilitation and recovery: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2016;47:e98–169 [DOI] [PubMed] [Google Scholar]
- 17.Flux E, van der Krogt MM, Cappa P, et al. : The human body model versus conventional gait models for kinematic gait analysis in children with cerebral palsy. Hum Mov Sci 2020;70: 102585. [DOI] [PubMed] [Google Scholar]
- 18.Tang A, Eng JJ, Rand D: Relationship between perceived and measured changes in walking after stroke. J Neurol Phys Ther 2012;36:115–21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wonsetler EC, Bowden MG: A systematic review of mechanisms of gait speed change post-stroke. Part 1: spatiotemporal parameters and asymmetry ratios. Top Stroke Rehabil 2017;24:435–46 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kim CM, Eng JJ: Magnitude and pattern of 3D kinematic and kinetic gait profiles in persons with stroke: relationship to walking speed. Gait Posture 2004;20:140–6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Marrocco S, Crosby LD, Jones IC, et al. : Knee loading patterns of the non-paretic and paretic legs during post-stroke gait. Gait Posture 2016;49:297–302 [DOI] [PubMed] [Google Scholar]
- 22.Beyaert C, Vasa R, Frykberg GE: Gait post-stroke: pathophysiology and rehabilitation strategies. Neurophysiol Clin 2015;45(4–5):335–55 [DOI] [PubMed] [Google Scholar]
- 23.Sloot LH, van der Krogt MM: Interpreting joint moments and powers in gait, in: Muller B, Wolf SI (eds): Handbook of Human Motion. Berlin, Germany, Springer International Publishing AG, 2018:625–43 [Google Scholar]
- 24.Chockalingam N, Healy A, Needham R: Interpreting ground reaction forces in gait, in: Muller B, Wolf SI (eds): Handbook of Human Motion. Berlin, Germany, Springer International Publishing AG, 2018:609–23 [Google Scholar]
- 25.Parvataneni K, Olney SJ, Brouwer B: Changes in muscle group work associated with changes in gait speed of persons with stroke. Clinical biomechanics 2007;22:813–20 [DOI] [PubMed] [Google Scholar]
- 26.Jonsdottir J, Ferrarin M: Gait disorders in persons after stroke, in: Muller B, Wolf SI (eds): Handbook of Human Motion. Berlin, Germany, Springer International Publishing AG, 2018:1205–16 [Google Scholar]
- 27.Wonsetler EC, Bowden MG: A systematic review of mechanisms of gait speed change post-stroke. Part 2: exercise capacity, muscle activation, kinetics, and kinematics. Top Stroke Rehabil 2017;24:394–403 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ambrosini E, De Marchis C, Pedrocchi A, et al. : Neuro-mechanics of recumbent leg cycling in post-acute stroke patients. Ann Biomed Eng 2016;44:3238–51 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ploughman M, Kelly LP: Four birds with one stone? Reparative, neuroplastic, cardiorespiratory, and metabolic benefits of aerobic exercise poststroke. Curr Opin Neurol 2016;29:684–92 [DOI] [PubMed] [Google Scholar]
- 30.Semaan MB, Wallard L, Ruiz V, et al. : Is treadmill walking biomechanically comparable to overground walking? A systematic review. Gait & posture 2022;92:249–57 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The full deidentified data set will be available upon request.