Participants
The study was approved by the Institutional Review Board of Nambu University (IRB: 1041478-2020-HR-031). Eight stroke patients participated; however, one patient in poor condition was removed during the study. Therefore, seven patients (sex male/female: 6/1; paralyzed side left/right: 3/4; age: 71.43 ± 9.64 years, body mass: 67.00 ± 9.47 kg, height: 170.71 ± 6.52 cm; onset period: 12.57 ± 3.10 month; mini mental state examination, MMSE: 24.86 ± 0.90 score; modified Ashworth scale, MAS: 1.21 ± 0.27 grade) completed the study. The inclusion criteria were a diagnosis of stroke over 6 months ago, ability to move ankles without assistance, and ankle MAS below Grade 2. The exclusion criteria were as follows: cardiovascular or respiratory disease, orthopedic diseases in the legs, vision or hearing disabilities, and skin diseases of the feet. All participants were informed of the benefits and risks of this study, and written informed consent was obtained from all participants.
Study procedures
This study was a crossover study. Assessments performed by each participant were assessed in a physical therapy room at the Department of Rehabilitation Medicine, Suwan Medical Center. Prior to the assessment, participants received instruction on how to perform VR and non-VR (NVR) exercises. During this orientation, participants were familiarized with the procedures, assessment tools, and VR equipment of the study. One day after the orientation session, each participant completed two assessment sessions in a randomized order, with 48 h of rest between each session. Before the assessment session, each participant completed general physical therapy and rehabilitation exercises. Participants rested before the assessment session. An assessment session was conducted in the afternoon. An assessor prepared randomly shuffled sticks (A stick: VR; and B stick: NVR) and sealed each stick in an opaque envelope. Each participant looked for an envelope and opened the envelope to identify the exercise of assignment. The identified exercise was performed first, and the other exercise was performed two days later. The paretic leg of the participants was assessed using ROM of ankle dorsiflexion and plantarflexion, muscle stiffness of the tibial anterior muscle, and medial and lateral gastrocnemius muscles. After the completion of the pre-test assessments, each participant performed the intervention. Each intervention was performed in three sessions for 1 minute per session, and a 30-second rest was taken between each session. Immediately after the intervention, post-test assessments were conducted in the same order as pre-test measures. Participants completed the intervention and its measurements on the same day. One participant in the poor condition was lost to follow-up for the second session. A flowchart of the experimental design is shown in Figure 1.
ROM of ankle
The ROM of the ankle was measured using a plastic goniometer. The angle of ankle dorsiflexion in the prone position with 90 ° of knee flexion was measured. The angle of ankle plantarflexion in the supine position and the ankles outside the bed were measured. The axis of the plastic goniometer was placed on the lateral malleolus. The fixed arm was placed parallel to the line connecting the fibular head, and the moving arm was placed parallel to the line connecting the metatarsal bone of the fifth toe [17]. The measurements were performed by a physical therapist who was blinded to the measured values. The measurer told the recorder when the measurement was complete. The recorder visually confirmed and recorded the measured value. The data used the average of the two measured values. The plastic goniometer showed high interrater reliability (ICC = 0.87) and intrarater reliability (ICC = 0.91) [17].
Muscle stiffness of ankle
The stiffness of the ankle muscles of the tibialis anterior, gastrocnemius lateral, and medial muscles was measured using the myotonPro (Myoton AS, Tallinn, Estonia). The stiffness of the tibialis anterior in the supine position was measured [18]. The stiffness of the gastrocnemius lateral and medial muscle was measured with the patient in the prone position and with feet hanging over the end of the bed [19]. The measurements were performed by a physical therapist who was blinded to the measured values. The measurer told the recorder when the measurement was complete. The recorder visually confirmed and recorded the measured value. The data used the average of the two measured values. The myotonPro showed high interrater reliability (ICC = 0.93) and intrarater reliability (ICC = 0.95) [20,21].
Exercise protocols
VR
Participants performed VR using a vibrating foam roller (Vyper, Hyperice, Irvine, CA, US). Participants positioned the vibrating foam roller below the gastrocnemius of their paretic side leg. The frequency of VR was 28 Hz, which has been used in many prior studies [13, 22]. Thereafter, patients performed 60 s of dorsiflexion and plantarflexion of their ankle (Fig. 2). The physical therapist observed and encouraged the patient to continuously move through the entire ROM. Patients engaged in 30 s of rest in between exercises. Each exercise was performed three times.
NVR
The exercise protocols were the same as those used for the VR exercise, except vibration (vibration button off).
Statistical analyses
All data analyses were performed using SPSS version 25 (Chicago, IL, USA). Data are presented as the mean ± standard deviation (SD). Data were not observed statistically for normality (Shapiro–Wilk’s test, p < 0.05), and a few variables were normally distributed. Therefore, nonparametric tests were used. Descriptive statistics were performed for the characteristics of the participants. A Mann-Whitney U test was performed to analyze the differences between VR and NVR by comparing the differences between pre- and post-treatment measurements. The Wilcoxon test was performed to compare pre-and post-intervention results in each group. The effect size (Cohen’s d), which is the difference between the pre- and post-means divided by their common SD, was calculated and interpreted as small (d = 0.2), medium (d = 0.5), or large (d = 0.8) to present the magnitude of the effect [23]. The significance level (α) was considered to be p < 0.05.