Secondary Logo

Journal Logo

Whole-Body Vibration Intensities in Chronic Stroke: A Randomized Controlled Trial


Medicine & Science in Sports & Exercise: July 2016 - Volume 48 - Issue 7 - p 1227–1238
doi: 10.1249/MSS.0000000000000909

Purpose A single-blinded randomized controlled study was conducted to investigate the effects of different whole-body vibration (WBV) intensities on body functions/structures, activity, and participation in individuals with stroke.

Methods Eighty-four individuals with chronic stroke (mean age = 61.2 yr, SD = 9.2) with mild to moderate motor impairment (Chedoke–McMaster Stroke Assessment lower limb motor score: median = 9 out of 14, interquartile range = 7–11.8) were randomly assigned to a low-intensity WBV, high-intensity WBV, or control group. The former two groups performed various leg exercises while receiving low-intensity and high-intensity WBV, respectively. Controls performed the same exercises without WBV. All individuals received 30 training sessions over an average period of 75.5 d (SD = 5.2). Outcome measurements included knee muscle strength (isokinetic dynamometry), knee and ankle joint spasticity (Modified Ashworth Scale), balance (Mini Balance Evaluation Systems Test), mobility (Timed-Up-and-Go test), walking endurance (6-Minute Walk Test), balance self-efficacy (Activities-specific Balance Confidence scale), participation in daily activities (Frenchay Activity Index), perceived environmental barriers to societal participation (Craig Hospital Inventory of Environmental Factors), and quality of life (Short-Form 12 Health Survey). Assessments were performed at baseline and postintervention.

Results Intention-to-treat analysis revealed a significant time effect for muscle strength, Timed-Up-and-Go distance, and oxygen consumption rate achieved during the 6-Minute Walk Test, the Mini Balance Evaluation Systems Test, the Activities-specific Balance Confidence scale, and the Short-Form 12 Health Survey physical composite score domain (P < 0.05). However, the time–group interaction was not significant for any of the outcome measures (P > 0.05).

Conclusion The addition of the 30-session WBV paradigm to the leg exercise protocol was no more effective in enhancing body functions/structures, activity, and participation than leg exercises alone in chronic stroke patients with mild to moderate motor impairments.

Supplemental digital content is available in the text.

1Department of Physiotherapy, Guangdong Provincial Work Injury Rehabilitation Hospital, Guangzhou, CHINA; 2Department of Rehabilitation Sciences, Hong Kong Polytechnic University, Hong Kong, CHINA; and 3School of Allied Health Sciences, Griffith University, Gold Coast, AUSTRALIA

Address for correspondence: Marco Y.C. Pang, Ph.D., Department of Rehabilitation Sciences, Hong Kong Polytechnic University, Hung Hom, Hong Kong, China; E-mail:

Submitted for publication September 2015.

Accepted for publication February 2016.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (

Stroke is one of the leading causes of long-term disability and is an important public health concern (9). One of the major physical impairments experienced by individuals with stroke is muscle weakness, particularly in the paretic limbs (8,15,32). It is well documented that isometric and dynamic muscle strength are correlated to other important functions, such as walking endurance (8,15), walking velocity (8,15), and balance skills (15,24). Limitations in these functional activities may lead to poor community reintegration, triggering a vicious cycle of further deterioration of physical functioning, and reduced societal participation (32). Thus, it is important to tackle problems arising from muscle weakness and associated functional issues in stroke rehabilitation.

Whole-body vibration (WBV) therapy, in which vibratory signals are delivered to the human body through a vibration platform, has gained increasing attention in neurorehabilitation. Only eight randomized controlled studies have examined the efficacy of WBV on different aspects of neuromotor function poststroke, such as leg muscle strength, balance, spasticity, and mobility (5,6,17,22,34–36,38), with inconsistent results observed. Of the seven studies that measured leg muscle strength (5,17,22,34–36,38), three (43%) reported significant effects associated with WBV (34–36). Two (29%) of the seven studies that measured balance (5,6,17,22,23,34,38) generated positive findings, with Chan et al. (6) reporting an improvement in weight distribution between the two legs in standing, whereas Tankisheva et al. (34) reported an increase in equilibrium score during standing on a moveable surface (34). In addition, a significant reduction of spasticity (6,26) and an improvement in mobility after WBV (6,23) were reported in 50% and 40% of the trials, respectively. A recent meta-analysis of eight WBV trials on patients with stroke revealed that the effects of WBV on muscle strength and mobility performance remain inconclusive (40). Another recent systematic review of nine WBV clinical trials on stroke patients generated a similar conclusion (18). The limited number of studies and their methodological weaknesses (only two of the trials provided level one evidence) could partially account for the inconclusive results (18). Thus, good-quality clinical trials with larger study cohorts are required to study the efficacy of WBV among individuals with stroke (18,40).

Discrepancies in results observed within the literature could be due to the vast differences in WBV parameters (WBV intensity, choice of exercise, and exercise duration) and characteristics of participants across studies (18). Indeed, specific factors required for a successful treatment outcome cannot be identified because differences exist in multiple parameters across studies. No previous trial has attempted to vary any of the above parameters and compare their effects. Among the various parameters, WBV intensity may be a key factor in determining treatment success. Previous EMG studies in able-bodied individuals (29) and individuals with stroke (19,21) revealed that the EMG amplitude of major leg muscle groups during WBV exposure was significantly augmented as WBV intensity was increased. Thus, it was postulated that higher WBV intensities should be more effective than lower intensities in improving leg muscle strength after a period of exercise training. The current study was designed in such a way that the effects of two different WBV intensities on leg muscle strength could be compared.

Another knowledge gap identified in systematic reviews is the relative lack of activity and participation outcomes in previous stroke WBV trials (18,40). The health consequences of stroke are multidimensional, and this context should be taken into consideration when selecting outcome measures. Therefore, the framework of this study was constructed based on the International Classification of Functioning, Disability, and Heath model (39), by incorporating outcomes of body functions/structures (e.g., muscle strength and spasticity), activity (e.g., mobility, walking endurance, and balance), and participation levels (e.g., participation in community activities). According to this model, there is a dynamic interaction between body structures/functions, activity, and participation, meaning that impairments in body structures/functions may influence activity and participation outcomes. Therefore, by addressing impairments (e.g., muscle weakness) through the proposed intervention, it was postulated that activity and participation may also improve. The inclusion of outcome measures in body structures/functions, activity, and participation domains would provide a more comprehensive picture of the therapeutic value associated with the experimental intervention.

The objective of this randomized controlled trial (RCT) was to investigate the effects of different WBV intensities on body functions/structures, activity, and participation in community-dwelling individuals with chronic stroke. It was hypothesized that 1) adding WBV to exercise training would lead to significantly greater improvements in body functions/structures, activity, and participation outcomes, compared with the same exercise regimen without WBV, and 2) the high-intensity protocol would induce significantly more gain in the same outcomes, compared with the low-intensity protocol.

Back to Top | Article Outline



The present investigation was a single-blinded RCT, in which the assessor was blinded. The study was registered at (NCT01822704). The reporting of this WBV clinical trial was in accordance with the recommendations of the International Society of Musculoskeletal and Neuronal Interactions (31).

Back to Top | Article Outline

Participants and Sample Size

The study was conducted at a research laboratory at the Hong Kong Polytechnic University. The inclusion criteria were as follows: diagnosis of hemispheric stroke persisting for more than 6 months before the time of enrolment, age ≥18 yr, community dweller, a score of 6 or higher on the Abbreviated Mental Test (1), and the ability to stand with or without aid for more than 90 s. Patients were excluded if any of the following conditions were present: brainstem or cerebellar stroke, other neurological disorders (e.g., spinal cord injury), neoplasms, severe cardiovascular diseases (e.g., requiring a pacemaker and uncontrolled hypertension), pain that affected the ability to participate in physical activities, pregnancy, vestibular conditions, recent fractures or metal implants in the lower limbs, or other serious medical problems.

The sample size was estimated based on evidence from a previous WBV study that investigated the leg extensor EMG activity during WBV in individuals with stroke (19), using G Power 3.1 software (Universitat Dusseldorf, Germany). Liao et al. (19) demonstrated that WBV training induced significantly higher levels of muscle activity in the paretic leg, with effect sizes (f) of 0.46–0.93 (i.e., large effect sizes). To be more conservative, a medium effect size was assumed (convention: f = 0.25). Based on a 2 × 3 repeated-measures ANOVA, with an alpha value of 1% and power of 80%, the minimum sample size required to detect a significant group–time interaction effect would be 21 subjects in each group (total of 63 participants). We used a more stringent alpha value because of the inflated probability of making type I errors due to multiple testing. To account for a 15% attrition rate, we aimed to recruit a minimum of 75 participants (25 per group). Written informed consent was obtained from all participants. The principles of the Declaration of Helsinki were followed, and the study was approved by the Human Research Ethics Review Subcommittee of the Hong Kong Polytechnic University.

Back to Top | Article Outline

Recruitment and Randomization

The recruitment of participants took place from February 2013 to February 2014 via the Hong Kong Stroke Association. Those who expressed interest in participating in the study were initially screened through telephone interviews, followed by a face-to-face assessment session. After eligibility was confirmed, participants were then randomized into the low-intensity WBV group (LWBV), the high-intensity WBV group (HWBV), or the control (CON) group using a 1:1:1 allocation ratio (Fig. 1). To ensure concealed allocation, subjects were randomly assigned to groups using sealed opaque envelopes distributed by an “off-site” researcher who was not involved in the recruitment of participants, provision of exercise training, or measurement of outcomes. The last participant completed the postintervention assessment on May 20, 2014.



Back to Top | Article Outline


All participants received WBV exercise training three times a week for a total of 30 sessions. A minimum 1-d rest period was scheduled between training sessions. Extra sessions for missed appointments were arranged to ensure that all participants completed all 30 sessions. All exercise training took place in the same research laboratory of the Hong Kong Polytechnic University. Exercise sessions for the three groups took place at different times of the day, such that the participants from each treatment group could not observe what exercises the other groups received. Each exercise session began with 10 min of warm-up exercises and ended with 10 min of cooldown exercises (general stretching exercises in a sitting position and exercise using a cycle ergometer).

Participants in the LWBV group (n = 28) and HWBV group (n = 28) received exercise training on a WBV platform that delivered synchronous WBV (Gymna Fitvibe Medical System, Gymna Uniphy Pasweg, Bilzen, Belgium). The choice of WBV and exercise protocols (Table 1) was adapted from a previous study that examined muscle activity during WBV exposure among individuals with stroke (19). In that study, WBV intensities similar to our LWBV protocol induced significantly higher leg muscle activity compared with the control condition (19). The highest level of leg muscle activity was attained during deep squat, semisquat, forward, and backward weight-shift exercises (19). Therefore, these exercises were chosen in this study to optimize the activation of major leg muscle groups (Table 1). Knee extension while in an erect standing posture was avoided to minimize the transmission of WBV to the head (29). In addition to dynamic exercises, static exercises were included in the training protocol because daily activities involve both static (isometric) and dynamic muscle work. Indeed, similar to dynamic muscle strength (8,24), isometric leg muscle strength has been shown to be strongly correlated with other important functions poststroke, including walking endurance, walking velocity, and balance ability (15). Moreover, previous WBV trials in older adults and individuals with stroke have provided no clear evidence that using a combination of static and dynamic exercises (23) is inferior to dynamic exercises (17,35) or static exercises alone (5,34,38).



Dynamic exercises (exercises 1–3, Table 1) were performed in cycles of 3 s, with 20 repetitions per minute. A metronome was used to pace the participants in performing the exercises at the desired rhythm. A rhythm of 20 repetitions per minute was selected, based on our pilot study, which demonstrated that most individuals were able to perform the exercises at this pace for 1.5 min without experiencing excessive fatigue, while finding it sufficiently challenging. For the static exercise (Exercise 4, Table 1), the participants were asked to sustain the semisquat position for 1.5 min in each repetition. A rest period of 1.5 min was given between each exercise repetition.

The training protocol had a progressive design, with a gradual increase in the duration of exercise (from 12 to 18 min per session) over the course of the treatment period, as tolerated. We increased the exercise duration as a means of progressing the intervention, given that reduced exercise endurance is often a problem poststroke (32). We did not increase the frequency of exercise training sessions (e.g., 5 d·wk−1), mainly due to feasibility issues. In the current training program, patients were already required to travel three times a week to our exercise facility, which was quite a distance from the homes of several participants. Furthermore, patients were often accompanied by their primary caregiver. Increasing the frequency of training may have increased travel expenses, burden to the caregiver, and possibly the attrition rate. Nonetheless, the exercises used in this study were quite challenging for the participants, who had varying degrees of physical impairments. Having at least one rest day after a training session was designed to facilitate participant recuperation and to enable better exercise performance in the next training session. The exercises were progressed only if tolerated by the participants. The RPE was also monitored (2). If the participant reported an RPE > 15, the exercise would be terminated, and a longer rest period (e.g., 3 min or longer, as requested by the participants) was given before proceeding to the next exercise. An RPE value of 15 was chosen because it was found to be the optimal cutoff for identifying individuals who attained a peak RER of ≥1.10 (an indicator of maximal exertion) (27). We used the same standard to monitor all exercises to ensure safety. Apart from the longer rest period, no modification of the rhythm or movement was made to the next exercise if an RPE > 15 was reported during the previous exercise. If the participant reported excessive fatigue, muscle soreness, or pain from the previous training session, the exercise duration was not progressed during that session.

The WBV settings were validated by a triaxial accelerometer (Model 7523A5; Dytran Instruments Inc., Chatsworth, CA), as recommended by the International Society of Musculoskeletal and Neuronal Interactions (31). The frequency of the WBV signals used was 20 Hz and 30 Hz. Frequencies higher than 30 Hz and amplitudes higher than 1 mm were not used in this study due to the very high peak acceleration values generated (14). Signal distortion is more severe with high-amplitude vibration signals (14). WBV frequencies below 20 Hz were not used because they may induce a considerable resonance effect, resulting in amplification of the vibration signals and possible adverse effects (14). Our pilot work also showed that the high-intensity protocol demanded a substantial exercise effort from the stroke participants, without causing excessive fatigue.

The CON group completed the same movements while standing on the same WBV platform, but no WBV was delivered (i.e., WBV device was turned off). Treatment sessions for all three groups were supervised by a researcher (researcher/participant ratio = 1:2). All participants performed the same exercises while standing on the WBV platform (Table 1). The training instructions and exercise progression pattern were the same for all three groups. The participants were asked to report to the research team if there was any change in medications during the study period.

Back to Top | Article Outline

Outcome Measures

Outcome measurements were performed between February 2013 and May 2014. Demographics and other relevant information (i.e., medications, medical history) were collected at the baseline assessment. The level of impairment of the leg and foot was evaluated using the Impairment Inventory of the Chedoke–McMaster Stroke Assessment (CMSA) (11). The rating for each body part was based on a seven-point ordinal scale, with higher scores indicating better motor recovery. The ratings for the leg and foot were summed to yield an overall CMSA motor score for the paretic lower limb. The Functional Ambulation Category (score range = 0–5; 0 = nonambulatory, 5 = independent) was used to indicate walking ability (13). The following outcomes were measured at baseline (within 1 wk before the commencement of the exercise training) and postintervention (within 1 wk after the completion of 30 treatment sessions) by the same blinded assessor.

Back to Top | Article Outline

Primary Outcome

Muscle strength

The knee extension and the flexion muscle strength of both the paretic and the nonparetic leg were measured by a dynamometer (NUMAC® NORMTM Testing & Rehabilitation System, Computer Sports Medicine, Inc., Stoughton, MA). Isometric, isokinetic concentric, and eccentric muscle strength were tested. After a practice trial, each participant performed a maximal voluntary isometric contraction of knee flexion and extension at two knee joint angles, 30° and 70° of knee flexion, respectively. The peak torque value (N·m) was registered. Isokinetic knee concentric and eccentric flexion/extension contractions through a range of movement between 70° and 10° of knee flexion at a fixed angular velocity of 60°·s−1 were also measured. An angular speed of 60°·s−1 was chosen, as this has been commonly used in previous stroke studies (5,17,26,34). Previously, it has been noted that a substantial proportion of individuals with stroke are unable to perform at higher angular velocities due to factors such as severe spasticity, which was also apparent in our pilot testing. The peak power value (W) was recorded. For all test conditions, three trials were performed with a 2-min rest period between trials. Data were then averaged and normalized by the participant’s body weight to yield the mean isometric strength (N·m·kg−1) and concentric and eccentric strength (W·kg−1) of knee flexion and extension in each leg. Muscle strength measurements using isokinetic dynamometry have been shown to be highly reliable in individuals with chronic stroke (5,17,26).

Back to Top | Article Outline

Secondary Outcomes


Spasticity in the knee extensors and ankle plantarflexors was assessed using the six-point Modified Ashworth Scale (MAS) (0 = no spasticity, 4 = affected part rigid). The MAS is widely used to evaluate muscle tone in stroke research and has acceptable reliability (Kendall’s tau correlation = 0.847) (25).

Back to Top | Article Outline


The 14-item Mini Balance Evaluation Systems Test (Mini-BESTest) was used to evaluate balance performance in everyday functional activities (37). The total score on this test ranges from 0 to 28, with higher scores indicating better balance ability. The Mini-BESTest has good psychometric properties when used in individuals with stroke, with excellent internal consistency (Cronbach’s alpha = 0.89–0.94), intrarater reliability (intraclass correlation coefficient [ICC] = 0.97), and interrater reliability (ICC = 0.96) (37).

Back to Top | Article Outline

Walking endurance

The 6-Minute Walk Test (6MWT) was administered while oxygen consumption (V˙O2) was continuously recorded using the FitMate™ metabolic system (Cosmed, Rome, Italy). At the beginning of each testing session, the system was calibrated according to the manufacturer’s guidelines. The total distance covered (m) and the mean V˙O2 rate (mL·kg−1·min−1) during the last 30 s of the 6MWT (a measure that is moderately associated with peak V˙O2 in individuals with stroke) were used for subsequent analysis (7). Both V˙O2 measured during the 6MWT and the distance covered have shown high test–retest reliability in individuals with stroke (ICC > 0.95) (7).

Back to Top | Article Outline

Functional mobility

Functional mobility was measured with the Timed-Up-and-Go (TUG) test (28). Each participant was asked to stand up from a chair, walk forward for 3 m, turn around, and walk back to the chair and sit down, as quickly as possible. The time taken to complete the test (in seconds) was measured using a stopwatch. The TUG test was carried out twice, with trials separated by a 1-min rest period. The average of the two trials was used for subsequent analysis.

Back to Top | Article Outline

Balance self-efficacy

Balance self-efficacy was evaluated using the Activities-specific Balance Confidence (ABC) scale (3). Participants were instructed to rate their level of confidence in performing each activity without losing their balance, using a numerical rating scale from 0 to 100, with higher scores denoting better balance confidence. The scores for each item were summed and then averaged to obtain the total ABC score. The ABC scale has been demonstrated to be a reliable and valid tool for evaluating balance self-efficacy in individuals with stroke (3).

Back to Top | Article Outline

Participation in daily activities

The Frenchay Activity Index (FAI) was used as a measure of participation (12). The FAI records the frequency of participating in social activities and performing more complex activities of daily living (e.g., domestic chores, outdoor mobility, and leisure). Each of the 15 items was rated on a scale from 0 to 3, yielding a total score of 15 to 60 (15–29: inactive or restricted participation; 30–44: active; and 45–60: highly active) (12). The construct validity and reliability of the FAI has previously been established (ICC = 0.87) (12).

Back to Top | Article Outline

Perceived environmental barriers to societal participation

Participants also rated their perception of environmental barriers to societal participation using the 25-item Craig Hospital Inventory of Environmental Factors (CHIEF) (20). The score for each of the 25 items was calculated by multiplying the magnitude score (small problem, 1; big problem, 2) by the frequency score (range: daily, 4; never, 0) to yield a product or overall “impact” score. Items relating to work or school, when the respondent was neither working nor in school, were considered “not applicable” and were not scored. The total CHIEF score is the mean of up to 25 overall impact scores. Liao et al. (20) demonstrated that the CHIEF is a reliable and valid tool for evaluating the perceived environmental barriers to societal participation among individuals with chronic stroke.

Back to Top | Article Outline

Quality of life

Quality of life was assessed using the Short-Form 12 Health Survey, version 2 (SF-12, Chinese version) (16). A mental health composite score and a physical composite score were generated (range: 0–100), with higher scores denoting better health-related quality of life.

Back to Top | Article Outline

Statistical Analysis

All statistical analyses were performed using IBM SPSS software (version 20.0; IBM, Armonk, NY, USA). A more stringent significance level of P < 0.01 was set due to the multitude of outcomes involved. Descriptive statistics (e.g., mean and standard deviation) were used to indicate central tendencies and variability of the data. One-way ANOVA (for continuous variables), chi-square tests (for nominal variables), and Kruskal–Wallis tests (for ordinal variables) were used to compare the baseline characteristics of the three groups. An intention-to-treat analysis was performed. For those who dropped out from the study, the results of the baseline assessment were carried over to the subsequent assessments using the last observation carried forward (LOCF) method (30). One key assumption of LOCF is that patients who do not receive treatment maintain status quo (30). However, in reality, many intervention programs are designed to prevent the deterioration of patients who are expected to get worse without intervention. Therefore, the LOCF method, by assuming that the past continues unchanged, may result in an overestimation of treatment efficacy or an underestimation of harmful effects. However, we chose to use LOCF because our participants were in the chronic stage of stroke and were all ambulatory and thus should not experience major deterioration in leg muscle strength or other health outcomes in the absence of the WBV intervention during the study period (about 75 d on average). Previous chronic stroke WBV trials also have not reported substantial deterioration in health outcomes in the CON group (5,17). Therefore, we felt that the LOCF is a reasonable imputation method, given the context of this study.

The Kolmogorov–Smirnov test was used to check normality of the data. ANOVA (mixed design; between-subject factor: group; within-subject factor: time) was used to compare outcome variables across the two time points (i.e., baseline and postintervention). Contrast analysis was performed within each group post hoc, when appropriate. As the MAS was an ordinal variable, between-group comparisons of the postintervention scores were made using the Kruskal–Wallis test (which generated H statistics that were tested using the chi-square distribution), followed by post hoc Mann–Whitney tests, as indicated. The above analyses based on the intention-to-treat (ITT) principle were repeated after excluding the dropouts (i.e., on-protocol analysis). Among the various outcome variables, only the 6MWT and the Mini-BESTest had well-established minimal clinically important difference (MCID) values, at 34.4 m (33) and 4 points (10), respectively. The proportion of individuals who achieved an improvement in the 6MWT ≥ 34.4 m or Mini-BESTest ≥ 4 points was compared across groups using the chi-square test.

Secondary analysis was performed to identify the factors that may be related to better treatment outcomes after WBV training. The change scores (postintervention score minus the preintervention score) of the LWBV and HWBV groups for each outcome were correlated with the corresponding baseline scores and relevant characteristics of the participants (e.g., age, time taken to finish the 30 sessions of exercise training, baseline outcome measure scores, etc.) using either Spearman’s rho or Pearson’s correlation, depending on whether the assumptions for parametric analysis were fulfilled.

Back to Top | Article Outline


Of 113 individuals with stroke who were screened for eligibility, 84 fulfilled all selection criteria (see the CONSORT flow diagram in Fig. 1). Twenty-eight participants were randomly allocated to each of the LWBV (8 women), HWBV (10 women), and CON (4 women) groups, respectively. A total of 74 participants completed the training programs and postintervention assessments (Fig. 1). The overall attrition rate was 11.9%. Of the four dropouts in the HWBV group, one withdrew after the baseline assessment, and three dropped out after having completed 1, 9, and 13 sessions, respectively. In the LWBV group, five individuals dropped out after having completed 3, 5, 6, 9, and 19 sessions, respectively. One participant from the CON group withdrew after having completed 24 sessions.

Back to Top | Article Outline


Demographic information is summarized in Table 2. All participants were ambulatory; 75 did not require any walking aid indoors. The CMSA motor score for the paretic lower limb (median = 9 out of 14, interquartile range = 7–11.8) revealed that the motor impairment level was mild to moderate. There was no significant between-group difference in any of the demographic (Table 2) or outcome variables at baseline (P > 0.05) (Tables 3 and 4). The on-protocol analysis after removal of dropouts yielded similar results (see Table, Supplemental Digital Content 1, which shows the on-protocol analysis, None of the participants reported any changes in medications throughout the study period.







Back to Top | Article Outline

Training Duration

Among those who completed all of the postintervention assessments, the mean number of days taken to complete the 30 sessions of exercise training showed no significant difference among the three groups (P = 0.729) (Table 2). The maximum time interval (mean number of days) between two training sessions was also similar among the three groups (P = 0.474) (Table 2).

Back to Top | Article Outline

Adverse Events

One participant from the LWBV group reported mild knee pain after WBV therapy and five reported fatigue (three from the LWBV group and two from the HWBV group) (Fig. 1). These participants eventually dropped out of the study. The remaining participants were able to increase their duration of exercise, as described in our protocol (Table 1).

Back to Top | Article Outline

Outcome Measures

ITT analysis

In the ITT analysis, there was a significant time effect for several muscle strength measures on the paretic side (i.e., isometric flexion and extension at 70°, and concentric flexion) (P < 0.01; Table 3), TUG, 6MWT distance, V˙O2 during 6MWT, Mini-BESTest, ABC, and the physical composite score domain of the SF-12 (P < 0.01; Table 4). However, none of the variables showed significant time–group interactions (P > 0.01). The Kruskal–Wallis test revealed no significant differences in the postintervention knee MAS (χ2 = 0.230, P = 0.891) or ankle MAS scores (χ2 = 0.642, P = 0.725) among the three groups. A total of 12, 11, and 12 individuals showed an improvement in the 6MWT ≥ 34.4 m (MCID value) in the CON, LWBV, and HWBV groups, respectively, after the intervention period. With respect to the Mini-BESTest scores, 18, 20, and 18 participants exhibited an increase of ≥4 points (MCID value) postintervention in the CON, LWBV, and HWBV groups, respectively. No significant between-group difference was found in the proportion of subjects who achieved an improvement in the 6MWT (χ2 = 0.098, P = 0.952) or Mini-BESTest (χ2 = 0.446, P = 0.800) that was at, or beyond, the respective MCID values. All analyses above were also performed after removal of the dropouts (i.e., on-protocol analysis), with similar results found (see Table, Supplemental Digital Content 1, which shows the on-protocol analysis,

Back to Top | Article Outline

Secondary Analysis

In the secondary analysis, an attempt was made to determine whether there was any significant association between the change score of each outcome measure and their respective baseline values and other relevant factors (e.g., training duration).

Back to Top | Article Outline

LWBV group

A significant negative correlation was found between the baseline scores and the change scores for knee flexion eccentric strength (r = −0.509, P = 0.006) in the paretic leg, indicating that the participants with poorer neuromuscular function tended to have greater improvement in this outcome.

Back to Top | Article Outline

HWBV group

There were significant negative correlations between the change scores and their respective baseline scores for concentric flexion (r = −0.510, P = 0.006) and extension strength (r = −0.832, P < 0.001), eccentric flexion (r = −0.554, P = 0.002) and extension strength (r = −0.554, P = 0.002) of the paretic knee, and isometric extension strength at 30° (r = −0.500, P = 0.007) and flexion concentric strength (r = −0.490, P = 0.008) of the nonparetic knee. As with the LWBV group, participants with poorer neuromuscular function tended to have greater improvements in these outcomes.

Back to Top | Article Outline


The key finding of this study was that the addition of LWBV or HWBV protocols to a leg exercise protocol was no more effective in enhancing body functions/structures, activity, and participation outcomes than leg exercises alone (i.e., the CON protocol).

Back to Top | Article Outline

Does WBV stimulation confer any additional benefits?

The findings of the current study were in contrast with both hypotheses established. In all cases, certain leg muscle strength variables showed significant time effects, indicating significant improvement after the training period. However, the group–time interaction effects were not significant, indicating that adding either LWBV or HWBV to the leg exercise protocol did not confer an additional therapeutic effect for strength outcomes.

Results related to the efficacy of WBV training on muscle strength in individuals with stroke after 4–12 wk of training have been mixed (18,40). Although several studies reported no significant effects (5,17,22,38), Tankisheva et al. (34) and Tihanyi et al. (35) observed positive effects with WBV on muscle strength outcomes. Although several reasons may exist for the discrepancies in results between the current report and these two studies, a key issue may be related to the design of the CON group. In both the current study and previous research that reported no significant between-group differences in muscle strength outcomes (5,17,22,38), the CON group performed exactly the same exercises as the WBV group. In contrast, Tankisheva et al. (34) and Tihanyi et al. (35) included a comparison group that engaged in different activities. Specifically, Tankisheva et al. (34) found a significantly greater increase in isometric and isokinetic knee extension torque (240°·s−1) in the paretic leg for the WBV group when compared with the comparison group that engaged in habitual physical activities. Thus, it is possible that better outcomes observed in the WBV group were related to the leg exercises performed while standing on the WBV device rather than the WBV stimulation itself. In the study by Tihanyi et al. (35), the WBV group (WBV plus conventional rehabilitation) experienced a significantly greater improvement in eccentric and isometric knee extension torque in both legs compared with a comparison group that received conventional rehabilitative treatment only. The positive improvement in muscle strength reported in the WBV group could be attributable to the leg exercises and longer total treatment time rather than the WBV stimulation. Considering the overall available evidence, no study has convincingly demonstrated that WBV stimulation has a positive effect on muscle strength in individuals with stroke. A recent meta-analysis also revealed that WBV induced no significant effect on isometric and eccentric knee extension strength among individuals with stroke (40). Therefore, the results of this study further consolidate the current body of evidence in showing that WBV had no effect on leg muscle strength poststroke.

An alternative explanation for the lack of significant effects on muscle strength is that the intensity of the WBV stimulation may not be high enough. Higher intensities were not used in this study as the high peak accelerations generated have the potential to cause injuries, such as damage to fragile bones and back pain (4,14). Increasing vibration acceleration magnitude has been found to be associated with an increased risk for developing greater lower back pain and disability over time among professional drivers, during a 12-month follow-up period (4). In repetitive exposure to high load, the possibility of fatigue damage to fragile bones cannot be excluded (14). On the basis of these reasons, although the duration of exposure in a typical WBV session is much shorter than that in occupational exposure and that the vibration-induced effects are short in duration (in the order of milliseconds), caution was exercised by avoiding higher WBV intensities in this study (14). A previous RCT has shown that WBV at an intensity of 0.96g–1.61g was not effective in improving muscle strength in people with chronic stroke (17,26). Adding WBV at an intensity of 1.61g has been shown to augment the EMG activity of the major leg muscle groups during exercise by approximately 10%–25% (19,21). The intensity used in this study for the HWBV group (3.62g) was even higher and would presumably have induced greater muscle responses. However, a recent study on individuals with chronic stroke (21) showed that the relationship between the WBV intensity and the level of muscle activation was nonlinear; increasing the WBV intensity from 0.96g to 1.61g only led to an additional 3%–5% increase in leg muscle EMG amplitude. Thus, further increasing the WBV intensity beyond a certain point may no longer effectively increase EMG activity. Therefore, despite the use of higher WBV intensities in this study, this may not necessarily translate into substantially higher muscle activation level compared with lower WBV intensities.

Finally, the lack of significant results in this study may be related to the observation that WBV may be beneficial for a very select group of individuals only. Secondary analysis revealed that those with more severe deficits had a tendency to gain a greater degree of improvement from WBV training.

Significant time effects were also detected for body functions/structures (V˙O2 rate), activity (TUG, 6MWT distance, and Mini-BESTest), and participation levels (ABC and physical health domain of SF-12), indicating that all three groups experienced improvement in these outcomes after the training period. However, the lack of a group–time interaction on these outcomes indicated that the WBV stimulation itself did not confer any additional effects. As WBV did not result in any significant effect on the muscle strength and spasticity variables (body functions/structures), a significant treatment effect on the related outcomes at the activity and participation levels, which often have multiple determinants, would not be expected. In addition, the fact that the WBV therapy did not involve any walking-related activities may account for the nonsignificant treatment effect on the mobility outcomes observed. Nevertheless, our results generally concurred with the available body of evidence. Only four studies have previously investigated the influence of WBV on mobility function poststroke after 3–8 wk of training (5,17,23,38), and only Merkert et al. (23) reported better performance in the TUG test in the WBV group. However, their CON group engaged in conventional rehabilitation, whereas the WBV group received additional WBV training on top of conventional rehabilitation. Thus, their WBV group had a greater total treatment time than the CON group. This factor, rather than WBV stimulation, may better explain the outcomes in the WBV group. Six studies have assessed balance after 3–12 wk of training (5,17,22,23,34,38), and only one of these reported positive results (34). Finally, only one study has investigated the effect of WBV therapy on social participation, using the Stroke Impact Scale, and found no significant effect compared with sham vibration (5). Taken together, the present results are generally in line with previous studies in showing that WBV does not induce improvement in activity and participation outcomes. A similar conclusion was made in a recent meta-analysis by Yang et al. (40).

Back to Top | Article Outline


This study has several limitations. First, the findings should not be generalized to patients who are in the acute/subacute stage of recovery, or who have severe motor impairments, because the participants in the present study were all in the chronic stage and had mild to moderate impairments poststroke. Second, participants and the trainer were not blinded to group allocation, but this would be difficult to achieve in the context of the exercise trials used. However, all efforts were made to minimize any possible bias (e.g., assessor blinding, separate exercise periods for the three groups, etc.). Third, we did not increase the frequency of training sessions as a means of progressing the exercises because of practical issues. Finally, no long-term follow-up assessment was performed. Any potentially beneficial or harmful long-term effects of WBV remain uncertain.

Back to Top | Article Outline

Future directions

On the basis the results of the current study and the body of evidence accumulated from previous trials (18,40), the clinical use of WBV in stroke rehabilitation could not be recommended at this point. Further research should be performed in this area. As revealed in our secondary analysis, WBV may be more applicable to those who are more severely impaired. Future research should test the feasibility and efficacy of using WBV in a more homogeneous sample of people who have a greater severity of stroke-induced motor impairment. It would be interesting to also assess the effects of WBV in acute stages of stroke recovery, when impairment levels are often more severe. Furthermore, we only compared the effects of different WBV intensities, and further research is required to compare different WBV parameters (type of vibration stimulus and duration of WBV). For example, there is some preliminary evidence from a meta-analysis that vertical (synchronous) WBV is more beneficial than side-alternating WBV in improving mobility poststroke (40). Further studies are also needed to investigate fundamental questions such as the transmissibility of WBV signals and how transmissibility varies with different WBV parameters and exercises performed. Finally, electrophysiological studies should be performed to address the physiological mechanisms associated with the application of WBV poststroke.

Back to Top | Article Outline


In summary, although WBV therapy is safe and feasible for individuals with chronic stroke, the addition of the WBV paradigm used here (LWBV and HWBV protocols) to a leg exercise protocol was no more effective in enhancing body functions/structures, activity, and participation than leg exercise training alone in community-dwelling individuals with mild to moderate chronic stroke impairments.

This study was supported by the General Research Fund provided by the Research Grants Council (PolyU 5245/11E). The WBV device was provided by SOOST Limited.

All authors declare no conflict of interest. The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Antonelli Incalzi R, Cesari M, Pedone C, Carosella L, Carbonin PU; Gruppo Italiano di Farmacovigilanza nell’Anziano. Construct validity of the abbreviated mental test in older medical inpatients. Dement Geriatr Cogn Disord. 2003;15(4):199–206.
2. Borg G. Perceived exertion as an indicator of somatic stress. Scand J Rehabil Med. 1970;2(2):92–8.
3. Botner EM, Miller WC, Eng JJ. Measurement properties of the Activities-specific Balance Confidence Scale among individuals with stroke. Disabil Rehabil. 2005;27(4):156–63.
4. Bovenzi M. Metrics of whole-body vibration and exposure–response relationship for low back pain in professional drivers: a prospective cohort study. Int Arch Occup Environ Health. 2009;82(7):893–917.
5. Brogårdh C, Flansbjer UB, Lexell J. No specific effect of whole-body vibration training in chronic stroke: a double-blind randomized controlled study. Arch Phys Med Rehabil. 2012;93(2):253–8.
6. Chan KS, Liu CW, Chen TW, Weng MC, Huang MH, Chen CH. Effects of a single session of whole body vibration on ankle plantarflexion spasticity and gait performance in patients with chronic stroke: a randomized controlled trial. Clin Rehabil. 2012;26(12):1087–95.
7. Eng JJ, Dawson AS, Chu KS. Submaximal exercise in persons with stroke: test–retest reliability and concurrent validity with maximal oxygen consumption. Arch Phys Med Rehabil. 2004;85(1):113–8.
8. Flansbjer UB, Downham D, Lexell J. Knee muscle strength, gait performance, and perceived participation after stroke. Arch Phys Med Rehabil. 2006;87(7):974–80.
9. Fuentes B, Tejedor ED. Stroke. The worldwide burden of stroke-a blurred photograph. Nat Rev Neurol. 2014;10(3):127–8.
10. Godi M, Franchignoni F, Caligari M, Giordano A, Turcato AM, Nardone A. Comparison of reliability, validity, and responsiveness of the mini-BESTest and Berg Balance Scale in patients with balance disorders. Phys Ther. 2013;93(2):158–67.
11. Gowland C, Stratford P, Ward M, et al. Measuring physical impairment and disability with the Chedoke–McMaster Stroke Assessment. Stroke. 1993;24(1):58–63.
12. Holbrook M, Skilbeck CE. An activities index for use with stroke patients. Age Ageing. 1983;12(2):166–70.
13. Holden MK, Gill KM, Magliozzi MR, Nathan J, Piehl-Baker L. Clinical gait assessment in the neurologically impaired. Reliability and meaningfulness. Phys Ther. 1984;64(1):35–40.
14. Kiiski J, Heinonen A, Järvinen TL, Kannus P, Sievänen H. Transmission of vertical whole body vibration to the human body. J Bone Miner Res. 2008;23(8):1318–25.
15. Kluding P, Gajewski B. Lower-extremity strength differences predict activity limitations in people with chronic stroke. Phys Ther. 2009;89(1):73–81.
16. Lam CL, Tse EY, Gandek B. Is the standard SF-12 health survey valid and equivalent for a Chinese population? Qual Life Res. 2005;14(2):539–47.
17. Lau RW, Yip SP, Pang MY. Whole-body vibration has no effect on neuromotor function and falls in chronic stroke. Med Sci Sports Exerc. 2012;44(8):1409–18.
18. Liao LR, Huang M, Lam FM, Pang MY. Effects of whole-body vibration therapy on body functions and structures, activity, and participation poststroke: a systematic review. Phys Ther. 2014;94(9):1232–51.
19. Liao LR, Lam FM, Pang MY, Jones AY, Ng GY. Leg muscle activity during whole-body vibration in individuals with chronic stroke. Med Sci Sports Exerc. 2014;46(3):537–45.
20. Liao LR, Lau RW, Pang MY. Measuring environmental barriers faced by individuals living with stroke: development and validation of the Chinese version of the Craig Hospital Inventory of Environmental Factors. J Rehabil Med. 2012;44(9):740–6.
21. Liao LR, Ng GYF, Jones AY, Chung RC, Pang MY. Effects of vibration intensity, exercise, and motor impairment on leg muscle activity induced by whole-body vibration in people with stroke. Phys Ther. 2015;95(12):1617–27.
22. Marín PJ, Ferrero CM, Menéndez H, Martín J, Herrero AJ. Effects of whole-body vibration on muscle architecture, muscle strength, and balance in stroke patients: a randomized controlled trial. Am J Phys Med Rehabil. 2013;92(10):881–8.
23. Merkert J, Butz S, Nieczaj R, Steinhagen-Thiessen E, Eckardt R. Combined whole body vibration and balance training using Vibrosphere®: improvement of trunk stability, muscle tone, and postural control in stroke patients during early geriatric rehabilitation. Z Gerontol Geriatr. 2011;44(4):256–61.
24. Moriello C, Finch L, Mayo NE. Relationship between muscle strength and functional walking capacity among people with stroke. J Rehabil Res Dev. 2011;48(3):267–75.
25. Pandyan AD, Johnson GR, Price CI, Curless RH, Barnes MP, Rodgers H. A review of the properties and limitations of the Ashworth and modified Ashworth Scales as measures of spasticity. Clin Rehabil. 1999;13(5):373–83.
26. Pang MY, Lau RW, Yip SP. The effects of whole-body vibration therapy on bone turnover, muscle strength, motor function, and spasticity in chronic stroke: a randomized controlled trial. Eur J Phys Rehabil Med. 2013;49(4):439–50.
27. Pinkstaff S, Peberdy MA, Kontos MC, Finucane S, Arena R. Quantifying exertion level during exercise stress testing using percentage of age-predicted maximal heart rate, rate pressure product, and perceived exertion. Mayo Clin Proc. 2010;85(12):1095–100.
28. Podsiadlo D, Richardson S. The timed “Up & Go”: a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991;39(2):142–8.
29. Pollock RD, Woledge RC, Mills KR, Martin FC, Newham DJ. Muscle activity and acceleration during whole body vibration: effect of frequency and amplitude. Clin Biomech (Bristol, Avon). 2010;25(8):840–6.
30. Portney LG, Watkins MP. Validity in experimental design. In: Portney LG, Watkins MP, editors. Foundations of Clinical Research: Applications to Practice. 3rd ed. Upper Saddle River (NJ): Pearson/Prentice Hall; 2009. pp. 167–70.
31. Rauch F, Sievanen H, Boonen S, et al. Reporting whole-body vibration intervention studies : recommendations of the International Society of Musculoskeletal and Neuronal Interactions. J Musculoskelet Neuronal Interact. 2010;10(3):193–8.
32. Saunders DH, Greig CA, Mead GE. Physical activity and exercise after stroke: review of multiple meaningful benefits. Stroke. 2014;45(12):3742–7.
33. Tang A, Eng JJ, Rand D. Relationship between perceived and measured changes in walking after stroke. J Neurol Phys Ther. 2012;36(3):115–21.
34. Tankisheva E, Bogaerts A, Boonen S, Feys H, Verschueren S. Effects of intensive whole-body vibration training on muscle strength and balance in adults with chronic stroke: a randomized controlled pilot study. Arch Phys Med Rehabil. 2014;95(3):439–46.
35. Tihanyi J, Di Giminiani R, Tihanyi T, Gyulai G, Trzaskoma L, Horváth M. Low resonance frequency vibration affects strength of paretic and non-paretic leg differently in patients with stroke. Acta Physiol Hung. 2010;97(2):172–82.
36. Tihanyi TK, Horváth M, Fazekas G, Hortobágyi T, Tihanyi J. One session of whole body vibration increases voluntary muscle strength transiently in patients with stroke. Clin Rehabil. 2007;21(9):782–93.
37. Tsang CSL, Liao LR, Chung RCK, Pang MY. Psychometric properties of the Mini-Balance Evaluation Systems Test (Mini-BESTest) in community-dwelling individuals with chronic stroke. Phys Ther. 2013;93(8):1102–15.
38. van Nes IJ, Latour H, Schils F, Meijer R, van Kuijk A, Geurts AC. Long-term effects of 6-week whole-body vibration on balance recovery and activities of daily living in the postacute phase of stroke: a randomized, controlled trial. Stroke. 2006;37(9):2331–5.
39. WHO. International Classification of Functioning, Disability and Health (ICF). Geneva: World Health Organization; 2001. p. 25
40. Yang X, Wang P, Liu C, He C, Reinhardt JD. The effect of whole body vibration on balance, gait performance and mobility in people with stroke: a systematic review and meta-analysis. Clin Rehabil. 2015;29(7):627–38.


Supplemental Digital Content

Back to Top | Article Outline
© 2016 American College of Sports Medicine