Effect of Functional Rehabilitation on Performance of the Star Excursion Balance Test Among Recreational Athletes With Chronic Ankle Instability: A Systematic Review

Objective To determine (1) the effectiveness of rehabilitation for chronic ankle instability as measured by the Star Excursion Balance Test (SEBT) and (2) the relative efficacy and the long-term effects of these rehabilitation interventions. Data Sources Ten electronic databases were searched (2009-2019). Study Selection Included articles were randomized controlled trials in English investigating recreational athletes aged ≥18 years with chronic ankle instability. At least 1 functional rehabilitation intervention had to be included and the SEBT test (or the modified version) used as an outcome measure. Data Extraction Two researchers (L.A., O.N.) extracted data regarding participant demographics; intervention characteristics; trial size; and results at baseline, postintervention, and at follow-up, where appropriate. Data Synthesis A systematic review and narrative synthesis was conducted. Methodological quality of included studies was assessed using the Cochrane Risk of Bias Tool and the van Tulder scale. The review was registered with PROSPERO (ID: 164468). Ten studies (n=368), 2 high-quality, 1 moderate-quality, and 7 low-quality, were included in the review. Interventions included balance training, strength training, vibration training, and mixed training. Results suggest that rehabilitation of chronic ankle instability that includes wobble board exercises (average percentage change: 14.3%) and hip strengthening exercises (average percentage change: 12.8%) are most effective. Few studies compared different types of rehabilitation for chronic ankle instability. However, improvements on the SEBT suggest that a rehabilitation program focusing on wobble board training and hip strengthening performed 3 times weekly for 4-6 weeks is the optimal rehabilitation program to improve dynamic postural control in recreational athletes with chronic ankle instability. Conclusions Few studies directly compared different rehabilitation interventions, and there was limited long-term follow-up; therefore, the relative efficacy of different rehabilitation programs remains unclear. However, it seems that rehabilitation of chronic ankle instability should include proprioceptive and strengthening exercises of relatively short duration.

Balance training improves an individual's ability to maintain center of gravity and posture by challenging the vestibular and musculoskeletal systems. 48 It has been reported that balance training can improve dynamic postural control among individuals with chronic ankle instability. 43,49,50 McKeon et al 43 conducted a high-quality study investigating the effects of a 4-week balance program among adults with chronic ankle instability and reported significant improvements in the intervention group for self-reported disability and postural control. 43 Strength training involves exerting force in an attempt to surmount resistance, leading to greater recruitment and stronger synchronization of muscle fibers, 48 which improves neuromuscular control and muscular development. [51][52][53][54][55][56] Smith et al 57 conducted a high-quality study investigating the effects of a 6-week strength program among individuals with chronic ankle instability. They reported significant improvements in evertor strength in the intervention group and concluded that an effective strength program should be challenging and multiplanar to improve strength and prepare the ankle for return to regular activity. 57 Whole-body vibration training (WBVT) involves mechanical oscillations transmitted from a vibration platform that alters joint mechanoreceptors, muscle spindles, power, and strength performances, 58 but there is limited research exploring WBVT for chronic ankle instability, Ray 59 conducted a moderate quality meta-analysis (n=4) comparing the effectiveness of WBVT to wobble board rehabilitation. These authors concluded that wobble board training was more effective for improving dynamic postural control in recreational athletes with chronic ankle instability.
In 2010, Webster and Gribble 60 systematically reviewed functional rehabilitation literature for chronic ankle instability (n=6 randomized controlled trials). Their findings suggested that all functional rehabilitation interventions significantly improved postural control. 60 However, they did not compare interventions for their relative efficacy, nor did they assess follow-up periods to determine the long-term effects. This current review provides an up-to-date review of the most recent literature (the last 10 years) exploring the optimal rehabilitation parameters, as measured by the SEBT, to assist clinicians with the conservative management of chronic ankle instability.
The aim of this review is to determine the effectiveness of functional rehabilitation for improving dynamic postural control, as measured by the SEBT, among recreational athletes with chronic ankle instability, with specific consideration for the relative efficacy and long-term effects of interventions.

Study selection
Full-text randomized control trials published in English were included. For the purposes of this review recreational athletes were self-reported or defined as completing at least 20 minutes of moderate to high intensity physical activity 3 times per week. 62 It is recognized, however, that the definition of recreational athletes varies across studies. Brown et al, 62 for example, defined recreational athletes as those who participate in at least 20 minutes of physical activity 2 times per week. However, Ray 59 defined recreational athletes as individuals participating in more than 1.5 hours of moderate to vigorous physical activity per week. A previous study by Sierra-Guzman et al 47 did not even define recreational athletes. Functional rehabilitation was defined as "dynamic, closed-kinetic-chain activity other than quiet standing." 37(p99) Included trials were required to report the SEBT or Y Balance Test reach performances. Both short-term (6-12wk) and long-term (12+wk) follow-up studies were included. Table 1 displays the eligibility criteria for the included studies. The primary outcome of interest was improvements in the SEBT performances, expressed as a percentage of change relative to preintervention.

Data extraction
Two researchers (L.A., O.N.) extracted data regarding participant demographics, intervention characteristics, trial size, baseline and postintervention results, and follow-up results where relevant.

Risk of bias assessment
Two independent reviewers (L.A., O.N.) assessed each study using the Cochrane Risk of Bias Tool 63 and the van Tulder scale. 64,65 As recommended in the Cochrane Handbook for Systematic Reviews of Interventions 65 the Cochrane Risk of Bias Tool was used to assess 5 issues associated with risk of bias: sequence generation, allocation concealment, blinding of personnel and outcome assessors, incomplete outcome data, selective reporting, and additional possible threats to validity not previously identified. The van Tulder scale was also included because it assesses both compliance and timing of outcome assessments. Any ambiguity was discussed and a consensus reached, and disagreements were resolved by further discussion with D.O. or J.M.

Data synthesis
The data synthesis was conducted following the recommended standards of performance outlined by Eden et al 66 : description of the methodological characteristics of selected trials; strengths and limitations of each trial; how the limitations may have influenced the results; the relationship between the study characteristics and reported findings; and the relevance of each trial to its population, control, and outcomes of interest (table 2). Because there are numerous interventions used in the management of chronic ankle instability, with substantial clinical heterogeneity between studies, a meta-analysis was not conducted.

Quantifying the magnitude of results
The minimal detectable change (MDC) values outlined by Munro et al 67 (table 3) were chosen because they include the complete SEBT rather than a subsection and are more conservative. The average percentage change was calculated from the reach distances reported in the studies and compared with the average MDC value for those reach directions.   Table 4 displays the study characteristics of the included articles.
Control groups included normal activity, 38,45,47 general activity with strength training, 40 bicycle workout, 37 conventional physiotherapy, 46 no intervention, 39,44 or an active comparator. 35,36 Two studies included a follow-up period, both short-term. 36,47 Risk of bias Tables 6 and 7 present the outcomes of the Cochrane Risk of Bias Tool and the van Tulder scale. The mean score of the van Tulder scale was 5.1 of 11. Five studies had a high risk of bias, 37,38,[44][45][46] potentially caused by inadequate participant      ESs were large for BTP (Hedges' g=0.7) and STP (Hedges' g=0.6) groups.
1. Researchers were not blinded to group allocation.
Balance programs included multidirectional hopping, [35][36][37] progressive single-limb activities, [35][36][37] wobble board exercises, 44 and single-limb stance on different surfaces. 40 Three studies implemented balance programs, 1 high quality 35 and 2 low quality because of lack of blinding of assessors 36 and inadequate allocation concealment. 44 Anguish and Sandrey 35 reported significant improvements in the SEBT for the progressive hop-to-stabilization and singlelimb balance programs yielding an average percentage change of 4.52% and 4.89%, respectively, exceeding the corresponding MDC value of 4.48% (see table 5), indicating that these programs are effective.
Burcal et al 36 reported significant improvements in the SEBT producing an average percentage change of 8.9%, again exceeding the corresponding MDC value of 4.48 %. These authors reported that balance training alone showed an equal magnitude of change as balance training with sensorytargeted ankle rehabilitation strategies (average percentage change: 8.89%).
Linens et al 44 reported significant improvements in the SEBT for the wobble board group, producing the largest  average percentage change of 14.3%, exceeding the corresponding MDC value of 5.63%, indicating that this program is effective. However, this is a low-quality study because of lack of allocation concealment; therefore, these results should be interpreted with caution. Strength training included resistance band exercises for the ankle, 37,38 resistance TheraBand exercises for the hip, 39 proprioceptive neuromuscular facilitation training, 37 and single-leg heel raises. 37 Two studies conducted a strength program; both were deemed low-quality because of inadequate allocation concealment and lack of blinding of assessors. 38,39 Hall et al 38 reported that the resistance band group showed no significant improvements in the Y Balance Test, yielding a percentage change of 4.72%, which does not exceed their MDC of 7.7%, suggesting that this program is of limited effectiveness.
Smith et al 39 reported significant improvements in the SEBT for their hip strengthening group yielding an average percentage change of 12.8%, substantially exceeding the corresponding MDC value of 4.48%, suggesting that this program is effective One study implemented a vibration program involving single-leg heel raises and single-leg squats on a vibration platform. 45 Cloak et al 45 reported significant differences in the SEBT for the WBVT group compared with the controls, producing an average percentage change of 9.64%, exceeding the corresponding MDC value of 6.4%, again suggesting that the program is effective. This study is low quality because of inadequate allocation concealment; therefore, results should be interpreted with caution.
Three studies adopted a mixed training intervention incorporating general strength and balance training, 40 elastic tubing exercises and conventional physiotherapy, 46 and balance and vibration training. 47 One study was deemed high quality, 40 1 moderate-quality, 47 and 1 low quality. 46 Cruz-Diaz et al 40 reported significant improvements in the SEBT for the combined training group producing an average percentage change of 5.24%, exceeding the corresponding MDC value, suggesting that this program is effective.
Melam et al 46 reported significant improvements in the SEBT for the mixed training group producing an average percentage change of 2.34%, which does not surpass the corresponding MDC value of 5.63%. Sierra-Guzm an et al 47 reported significant improvements in the combined training group and the balance only training group, producing an average percentage change of 3.5% and 5.28%, respectively, which does not meet the corresponding MDC value of 5.63%, indicating that these programs are not effective.
One study compared balance and strength training 37 and reported large effect sizes for both groups, with the balance group displaying slightly greater effects than the strength training group. However, this study was low quality because of inadequate allocation concealment; therefore, results should be interpreted cautiously.

Intervention duration
Five studies involved a 4-week intervention, 35,36,39,44,46 and 5 studies implemented a 6-week intervention. 37,38,40,45,47 The average percentage change of the studies that included a successful 4-week 35,36,39,44 and 6-week intervention 37,40,45 was 13.4% and 7.44%, respectively. Hall et al 37 did not provide data to calculate the average percentage and therefore is not comparable. These results suggest that 4 weeks of rehabilitation intervention is a sufficient duration to produce results that are clinically significant.

Session frequency
One study completed 1 session per week, 45 8 completed 3 sessions per week, [35][36][37][38][39][40]44,47 and 1 completed 4 sessions per week. 46 Successful interventions included a frequency of 1 session 45 and 3 sessions per week, [35][36][37]39,40,44 with an average percentage change of 9.64% and 11.49%, respectively. Hall et al 37 did not provide data to calculate the average percentage and therefore is not comparable. These results suggest th\at 3 weekly sessions are sufficient to produce results that are clinically meaningful.

Discussion
This review aimed to determine (1) the effectiveness of rehabilitation for chronic ankle instability as measured by the SEBT and (2) the relative efficacy and the long-term effects of these rehabilitation interventions. The results suggest that rehabilitation of chronic ankle instability that includes wobble board exercises (average percentage change: 14.3%) 44 and hip strengthening exercises (average percentage change: 12.8%) 39 is the most effective because of a larger magnitude of change reported.
The benefits of wobble board rehabilitation for chronic ankle instability have been well-documented. [68][69][70][71] Strom et al 72 investigated peroneal muscle activity and frontal plane ankle kinematics during a single-leg stance on different surfaces. They reported that the wobble board produced the largest improvements in neuromuscular abilities and ankle sensorimotor control. 72 Research emphasizes that rehabilitation programs for chronic ankle instability should consider including wobble board exercises. [68][69][70][71]73,74 Previous research has reported that those with chronic ankle instability rely more on the hip's contribution during postural control tasks. 75,76 Individuals with chronic ankle instability display insufficiencies in hip external rotators and gluteus medius function. [77][78][79][80] Therefore, highlighting that rehabilitation programs should consider including hip strengthening exercises.
The suggested optimal frequency is 3 sessions per week (average percentage change: 11.49%). This is supported by 2 high-quality 35,40 and 4 low-quality studies. 36,37,39,44 The suggested optimal duration is 4 weeks (average percentage change: 13.4%). This result is supported by only 1 high-quality 35 and 3 low-quality studies. 36,39,44 Similarly, Powden et al 81 reported that improvements observed in individuals with chronic ankle instability after a 4-week multimodal intervention were equal to that of a 6-week intervention.

Relative efficacy of rehabilitation types
Only 1 study compared different rehabilitation types 37 and reported that balance training was slightly more effective. However, because of this study being low quality and the lack of studies that compared different rehabilitation types, these results are not conclusive.

Long-term effects
Burcal et al 36 reported that improvements were maintained at the 1-week follow-up displaying moderate to large effect sizes. Whereas Sierra-Guzm an et al 47 reported decreases at the 6-week follow-up with the nonvibration group displaying better ability to maintain postintervention improvements. Because of the lack of long-term follow-up assessments, the long-term effects of the interventions are unknown.

Comparison to previous literature
This is the only review in the last 10 years that has assessed the effectiveness of different rehabilitation types and suggested optimal rehabilitation parameters. Before this review, Webster and Gribble 60 investigated functional rehabilitation interventions for chronic ankle instability published from 1988-2008. They analyzed postural control outcome measurements in 6 studies, 60 reporting that a 4-to 6-week intervention with 3-5 weekly sessions can improve dynamic postural control. Webster and Gribble, 60 similar to this review, found wobble board rehabilitation effective for several stages of ankle instability. 60 Unlike this review, Webster and Gribble 60 assessed methodological quality of their studies using the Physiotherapy Evidence Database scale. However, the Physiotherapy Evidence Database scale has many short comings; it assesses the quality of reporting instead of characteristics that affect the risk of bias (which is recommended 63,82 ) and does not account for compliance or timing of outcomes, which are important when evaluating exercise interventions.

Study limitations
The importance of postural control is accepted for many clinical populations 83 ; however, the population of interest in this review was recreational athletes, and therefore the results may not be applicable to more general clinical groups. Second, the accumulated number of participants assessed is relatively small; including a study with a small sample size could have significantly influenced the magnitude of change between pre-and postintervention scores. Third, this review only analyzed dynamic postural control; by incorporating self-reported measures this may have provided a more indepth functional rehabilitation program for chronic ankle instability. Lastly, because a meta-analysis was not conducted, the findings of this review can only suggest optimal rehabilitation parameters; they are not conclusive.
Despite these limitations, this review rigorously evaluated risk of bias within and across the included studies. Furthermore, this is the only review that discusses an optimal rehabilitation program for recreational athletes with chronic ankle instability, thus assisting clinicians regarding the conservative management of chronic ankle instability.

Recommendations for future research
Future trials should be adequately powered and focus on meeting the minimum standards to reduce potential threats to bias. There is a need for trials to directly compare different rehabilitation types to provide a definite conclusion regarding the relative efficacy. Future trials should include a sufficient follow-up period to determine the long-term effects of an intervention.

Clinical relevance
This review suggests the optimal rehabilitation parameters required in the management of recreational athletes with chronic ankle instability. Three weekly sessions focusing on wobble board exercises and hip strengthening for 4-6 weeks is suggested. However, the evidence is insufficient for these results to be conclusive and are only suggestions to help guide clinicians in the management of chronic ankle instability.

Conclusions
Chronic ankle instability is associated with impaired sensorimotor control, which contributes to deficits in postural control activities. A rehabilitation approach focusing on wobble board exercises and hip strengthening performed 3 times weekly for 4-6 weeks is suggested to help improve dynamic postural control in recreational athletes with chronic ankle instability, at least in the short-term. The lack of long-term follow-up studies prevents definitive conclusions, and the results are suggested as a guideline to assist clinicians in the management of recreational athletes with chronic ankle instability. The long-term effects of the interventions remain unclear and further research is required.