Article Text

The assessment of postural control and the influence of a secondary task in people with anterior cruciate ligament reconstructed knees using a Nintendo Wii Balance Board
  1. Brooke E Howells1,2,
  2. Ross A Clark3,
  3. Clare L Ardern1,2,
  4. Adam L Bryant3,
  5. Julian A Feller1,4,
  6. Timothy S Whitehead4,
  7. Kate E Webster1
  1. 1Musculoskeletal Research Centre, Faculty of Health Sciences, La Trobe University, Melbourne, Victoria, Australia
  2. 2Department of Physiotherapy, La Trobe University, Melbourne, Victoria, Australia
  3. 3Department of Physiotherapy, The University of Melbourne, Melbourne, Victoria, Australia
  4. 4Epworth HealthCare, Richmond, Australia
  1. Correspondence to : Dr Kate Webster, Musculoskeletal Research Centre, Faculty of Health Sciences, La Trobe University, Bundoora, VIC 3083, Australia; k.webster{at}


Background Postural control impairments may persist following anterior cruciate ligament (ACL) reconstruction. The effect of a secondary task on postural control has, however, not been determined. The purpose of this case–control study was to compare postural control in patients following ACL reconstruction with healthy individuals with and without a secondary task.

Participants 45 patients (30 men and 15 women) participated at least 6 months following primary ACL reconstruction surgery. Participants were individually matched by age, gender and sports activity to healthy controls.

Materials Postural control was measured using a Nintendo Wii Balance Board and customised software during static single-leg stance and with the addition of a secondary task. The secondary task required participants to match the movement of an oscillating marker by adducting and abducting their arm.

Main outcome measures Centre of pressure (CoP) path length in both medial-lateral and anterior–posterior directions, and CoP total path length.

Results When compared with the control group, the anterior–posterior path length significantly increased in the ACL reconstruction patients' operated (12.3%, p=0.02) and non-operated limbs (12.8%, p=0.02) for the single-task condition, and the non-operated limb (11.5%, p=0.006) for the secondary task condition. The addition of a secondary task significantly increased CoP path lengths in all measures (p<0.001), although the magnitude of the increase was similar in both the ACL reconstruction and control groups.

Discussion ACL reconstruction patients showed a reduced ability in both limbs to control the movement of the body in the anterior–posterior direction. The secondary task affected postural control by comparable amounts in patients after ACL reconstruction and healthy controls. Devices for the objective measurement of postural control, such as the one used in this study, may help clinicians to more accurately identify patients with deficits who may benefit from targeted neuromuscular training programs.

Statistics from


Deficits in knee joint proprioception attributable to anterior cruciate ligament (ACL) injury have long been reported.1–3 However, the extent to which sensory deficits recover following ACL reconstruction surgery is yet to be fully understood. Although it is known that ACL reconstruction restores the mechanical stability of the knee,4 many individuals still fail to achieve optimal functional recovery5–8 and this may be linked with persistent sensorimotor deficits.4 ,9 Mechanoreceptors have been found in the native ACL and are thought to play an important role in knee joint proprioception and postural control.10 ,11 It has been suggested that ACL mechanoreceptors may not repopulate following surgery, resulting in altered afferent feedback pathways.12 Postural control relies on these peripheral proprioceptive pathways as the regulators of muscle tone.13–16

Most of the literature evaluating postural control following ACL reconstruction surgery has evaluated postural control in static single limb stance, where the participant's sole attention is on the postural task. A recent systematic review17 concluded that there were subtle, yet significant, impairments in postural control following ACL reconstruction when compared with healthy controls. However, the review also noted that it was difficult to draw firm conclusions from the body of current research due to the small number of studies and methodological differences.17

To further elucidate the postural control deficits occurring in the ACL-reconstructed population, a task that has greater cognitive and/or physical demands may be required. One way to achieve this could be with the use of a secondary task, which may better represent the many simultaneous physical and attentional demands required during daily living and on the sporting field. It has been established that central attentional resources are required to maintain an upright posture, and that the amount of resources required varies with the quality of the sensory information available.18 If sensory afferent information is affected after ACL reconstruction, through lack of mechanoreceptors or lack of adaptation of other tissues, increased central attentional capacity may be required to maintain balance. This has been shown in an electroencephalographic study in which greater activity was shown in the somatosensory parietal cortex in the ACL reconstructed group compared with controls.19

There is currently no research examining the effects of a secondary task on postural control in patients following ACL reconstruction. The purpose of this study was to compare postural control, with and without a secondary task that required controlled movement of the upper limb, in patients following ACL reconstruction with a control group. This task was chosen as many sports involve concurrent lower and upper body control. It was hypothesised that the ACL reconstruction group would display impaired postural control when compared with a control group, and that impairments in postural control would be more pronounced with the addition of a secondary task.



Sample size calculations were based on the premise of obtaining a moderate effect size,20 which required a minimum 68 participants (34 per group with β=0.2, α=0.05).

ACL reconstruction group

The ACL reconstruction group consisted of a sample of 15 women and 30 men who had undergone uncomplicated primary ACL reconstruction with a four-strand hamstring autograft. Participants were required to be at least 6-month postsurgery and were recruited January–March 2011 from the private practices of two experienced knee surgeons (JF and TW). Exclusion criteria were; (1) previous or concurrent posterior cruciate ligament injury, (2) previous or concurrent collateral ligament repair or reconstruction, (3) history of visual, vestibular or neurological disorders, (4) use of ototoxic medications (eg, chemotherapeutic agents such as methotrexate) likely to influence the postural control system within the previous 6 months, (5) greater than 2 years between ACL injury and ACL reconstruction and (6) history of lower limb (including ankle), neck or back injury in previous 6 weeks significantly affecting activity.

All participants underwent an arthroscopically assisted ACL reconstruction and were discharged from hospital on the first postoperative day. Postoperatively all patients undertook a rehabilitation protocol that encouraged immediate full knee extension and the restoration of quadriceps function as soon as possible. Particular emphasis was placed on the restoration of vastus medialis function. Weight bearing was allowed on an as-tolerated basis from the first postoperative day. No braces or splints were used. In general, most patients were full weight bearing by 2–3 weeks, riding a stationary bicycle by 4 weeks and running by 3–4 months. Sports-specific drills were introduced from 4 months.

Control group

The control group consisted of 15 women and 30 men, who were university staff and students recruited via advertisement and recreational athletes recruited via sporting clubs. Control participants had no history of ACL injury or surgery, and no lower limb injury within the previous 6 months that required surgery or restricted participation in activities of daily living for greater than 2 weeks. Other exclusion criteria were the same as for the ACL reconstruction group.

Matching of participants

Each ACL reconstruction participant was individually matched with a control participant by age (±4 years), gender and physical activity level. Physical activity level was determined using the Cincinnati Sports Activity Scale21 which categorises participants according to frequency of participation.

Materials and apparatus

Participants stood in a standardised position (using a grid) on a Wii Balance Board (WBB) (Nintendo, Japan) for testing of all postural control outcomes. The WBB has good-to-excellent test retest reliability (ICC=0.66–0.94) and excellent concurrent validity (ICC=0.77–0.89) compared to a laboratory-grade force platform.22 The WBB was placed on a hard surface (60cm×90cm×3 cm ply-wood) and calibrated by placing known weights on different positions of the WBB. The WBB was connected to a laptop computer via a Bluetooth connection and interfaced using custom LabVIEW 8.5 software (National Instruments, Austin, Texas, USA) written by one of the investigators (RC). An accelerometer in the Wii Remote (Nintendo) was used to provide the data for the arm movement and was connected to the computer via Bluetooth. All data were streamed to the software programme and displayed as soon as they were received, resulting in a typical feedback refresh time of less than 20 ms.


Ethics approval was obtained from the Faculty Human Ethics Committee and all participants provided written consent.

Task 1: single leg stance

Participants stood bare-footed on the WBB on one leg, with the middle of the longitudinal axis of their foot aligned with a line showing the centre of the WBB in the AP plane. They were instructed to stand as still as possible for 30 s (figure 1), with both limbs tested independently. Specific positioning regarding the weight-bearing knee (slightly flexed to approximately 20°), non-weight-bearing knee (flexed to 90°), non-weight-bearing hip (neutral flexion/extension) and hands (on hips) was adjusted on the basis of visual observation by the investigator. Participants fixed their gaze on a white dot displayed on a computer monitor positioned at eye-level 1.4 m from the WBB.

Figure 1

Task 1: participant position for single limb stance.

Task 2: single leg stance+secondary task

The protocol from task 1 was repeated with the addition of a secondary task. Participants were asked to stand on one leg, hold an accelerometer in the ipsilateral hand and repeatedly adduct and abduct the shoulder through full range of motion with the elbow extended (figure 2). The movement of the accelerometer resulted in the movement of a red marker on the computer screen (figure 3). Participants were instructed to match this red ‘angle’ marker to the movement of a larger yellow ‘target’ marker that moved vertically up and down on a sliding scale. This target marker was controlled by a sine wave oscillating at 0.33 Hz, resulting in 10 full adduction/abduction cycles per 30 s trial. Participants were reminded that the aim of the test was to ‘stand as still as possible’. This task was initially practised in bilateral stance to familiarise the participant to the sensitivity of the accelerometer. One practice trial in single leg stance was also completed.

Figure 2

Task 2: participant position for single limb stance+secondary task.

Figure 3

Participant view for single-limb stance and secondary task condition. During all trials centre of pressure feedback was not provided. The vertical bar on the right provided the arm adduction/abduction angle feedback via a small angle marker, which was matched to another larger, target marker that oscillated at 0.33 Hz. Reaching the top and bottom of the vertical bar corresponded to full range of motion arm abduction and adduction, respectively. Access the article online to view this figure in colour.

The order of completion of the postural control tasks, and the test limb, were randomised (via a coin toss) to minimise series effects. Participants completed one practice trial, and three trials of 30 s duration for both tasks to minimise the learning effects.23 There was a 30 s rest between trials, to minimise the effects of fatigue.24 Trials were excluded if the participant failed to remain upright for the trial duration.

Outcome measures

The primary outcome measures were centre of pressure (CoP) path length in the medial-lateral (ML) and anterior–posterior (AP) directions, and total CoP path length. CoP path length has been shown to be a reliable measure of postural control with movement of the CoP reflecting the neuromuscular control of the lower-limb musculature.21 CoP has been utilised by studies assessing postural control following ACL reconstruction.21 ,23 ,25

Data analysis

The data recorded from the WBB were interpolated to 100 Hz before being filtered using an undecimated wavelet-based filter (Symlet-8) with a low pass frequency of 12.5 Hz. All CoP path length data were exported to an Excel (Microsoft, Redmond Washington, USA) spreadsheet for analysis. Descriptive statistics were calculated for all demographic characteristics. Means and standard deviations were calculated for each of the dependent variables in both postural control tasks. The average of the three trials was used for statistical analysis. Between-group comparisons were made between the patient's operated limb and the corresponding side of the matched control.

After checking data for normality, a two-factor repeated measures analysis of variance (ANOVA) model was used to compare between the ACL reconstruction and control groups. This model included task (ie, task 1 and task 2) as a repeated factor and group as a between-subjects factor. For the between subject factor of group, separate ANOVAs were run where (1) the ACL reconstructed limb was compared to the control group and (2) the non-operated limb of the ACL patients was compared with the control group. For each ANOVA that resulted in a significant F ratio of p≤0.05, post hoc analysis was performed with t tests with the α-level adjusted to p≤0.025 for multiple comparisons. Paired t tests were also used to compare between the operated and non-operated limbs of the ACL reconstruction group at an α-level of p≤0.05. Analysis was performed using the Statistical Package for the Social Sciences (V.19, SPSS, Science Inc)


Demographic data for both groups are shown in table 1. Nineteen patients underwent right knee reconstruction and 26 patients left knee reconstruction. Nine patients underwent concurrent meniscal repair (five medial and four lateral) and one had medial meniscal resection. There were no significant differences in height or weight between the groups.

Table 1

Demographic data for ACL reconstruction and control groups

Single leg stance

Table 2 presents the mean and SD of CoP path length measurements.

Table 2

Anterior cruciate ligament reconstruction group (operated and non-operated) limb versus matched control limb: centre of pressure path length

AP CoP path lengths were greater in both the reconstructed and non-operated limbs of the ACL group compared with the control group for the single leg task. The difference was a significant 12.3% and 12.8% for the operated (p=0.02) and non-operated (p=0.02) limbs, respectively. There were no significant between group differences for the ML or total CoP path lengths or between limb differences for the ACL group.

Single leg stance+secondary task

As can be seen in table 2, the addition of the secondary task significantly increased all three CoP measures in both limbs of the ACL group as well as the control group (p<0.001). There were no significant task-by-group interactions. Mean CoP path length values were greater in both limbs of the ACL group compared with the controls; however, the only significant differences were between the non-operated limb of the ACL group compared to the controls for the AP direction (p=0.006) as well as total path length (p=0.015). There was no significant between-group difference for ML path lengths, nor were there any significant between-limb differences in the ACL group.


This study showed that the ACL-reconstructed group exhibited significantly greater excursion of AP CoP path lengths compared with the control group. The addition of the secondary task significantly increased CoP path length in both groups; however, the magnitude of change was similar.

The difference between the groups could be important given that a larger excursion of the CoP path may suggest a reduced ability to control the movement of the body during a static task, and thus suggest impairment in postural control. Previous studies21 ,26 ,27 have shown impairments in postural control in patients following ACL reconstruction when compared with controls, based on CoP measurements. In this study both the operated and non-operated limbs differed from the control group, which may suggest that the ACL reconstructed group was impaired bilaterally.

Previous research suggests that this is due to postural control being centrally mediated28–30 meaning that changes in the sensorimotor feedback from the operated limb also affect postural control during standing on the contralateral limb. Friden et al31 suggest that when ACL mechanoreceptors are stimulated on one side, this results in reflex γ-motor activity in both limbs. The results of this study may confirm such theories and suggest that sensory information from the ACL is important for bilateral postural control. A functional MRI study32 demonstrated neuroplastic changes in an ACL-deficient population compared with controls suggesting central changes may occur in response to the initial asymmetry, which may manifest as bilateral impairments in postural control.

Between-group differences were greatest in the AP direction. A possible explanation for this may be due to the fact that this is also the plane in which the muscle groups (ie, quadriceps and hamstrings) most affected by the injury and subsequent surgery function. The hamstrings are thought to be essential muscles for control of posture33 and reflex hamstring activity may be affected due to the loss of mechanoreceptors after ACL reconstruction, their inability to repopulate within the ACL graft or to the mechanical trauma occurring to the hamstring as part of the graft harvesting process. Studies have shown that there is a reflexive hamstring muscle response with stimulation of the ACL, suggesting it provides vital sensorimotor information for postural control.16 ,34 Consequently, the compromised proprioceptive and mechanical properties of the hamstrings after ACL injury and reconstruction may have contributed to the decreased postural control in the AP plane in this study. This AP dysfunction could also have an effect on the contralateral limb secondary to the theory noted above by Friden et al,31 in which a reduction in hamstring stretch receptor activity results in decreased reflex γ-motor activity and a consequent postural control impairment in the contralateral limb.

During testing it was evident that some patients had difficulty maintaining 90° knee flexion with their operated limb while standing on the non-operated limb, and this may have contributed to some of the increased movement about the non-operated limb. The most novel aspect of this study was the secondary task and it was clear that this made the postural control task more difficult, as both groups showed significant increases in postural control.35 However, despite the observation of a significant difference between the non-operated limb of the ACLR group and the control group, no differences in the magnitude of change in CoP values were observed between groups. For example, the percentage difference in AP CoP path length between the control limb and the operated and non-operated limbs for the secondary task trials were 8.4% and 11.4%, respectively. These magnitudes are actually lower than the 12.3% and 12.8% differences observed for the respective limbs during the standing single-limb balance trial.

This indicates that the additional demands of the secondary task did not place additional burden on the ACL reconstructed group. This in turn may suggest that the deficit in postural control after ACL reconstruction is a result of a compromised spinal reflex arc and/or muscle stretch reflexes, and is not related to changes in the higher central motor centres of the nervous system. Alternatively, the tasks used in this study may not have required the level of attentional resources needed to differentiate between two high-functioning groups. To further elucidate this mechanism, future research should assess the influence of tasks that require differing levels of physical and cognitive complexity. For example, a perturbation that is directed in the AP plane of movement to specifically target the postural deficit observed in this study and by Henriksson et al23 could be considered.

The strength of the study is individual matching of ACL reconstruction and control subjects. As many patients wish to return to sport following ACL reconstruction surgery, it was important that the control group was matched for physical activity. The clinical significance of our finding of impaired AP postural control must also be considered. Paterno et al36 observed that a 6-week neuromuscular training programme specifically designed to reduce ACL injury risk resulted in significantly improved AP but not ML postural control during a single leg balance test. This provides some evidence that the AP movement during a single leg postural control test may give an indication of an individual's lower limb neuromuscular function, and future studies should assess the relationship between AP postural control and functional performance throughout the ACL reconstruction rehabilitation process. If a relationship is observed between these measures, interventions such as those of Paterno et al36 which are known to improve balance deficits may be warranted. Balance board devices, such as the one used in this study, may represent a practical and inexpensive objective measure to help the clinician identify patients who have deficits in postural control.

A limitation of the current study was that the balance assessment task was not dynamic, and given that it was performed with the knee in a slightly flexed position of 20°, it did not place great demands on the control of valgus movement at the knee. This could be why we observed significant differences in the AP but not the ML direction, and future research should consider using either dynamic balance tasks or a more flexed knee position (ie, 60°) during static balance assessments to determine if ML plane balance deficits at the knee exist in this population. A further limitation was that data were not recorded for the accuracy of the secondary task, which was due to the priority of the programme being to record the postural control data and only using the real-time arm position data for feedback purposes. It is possible performance may have deteriorated in the secondary task rather than the postural control task based on conflicting evidence surrounding the central nervous system prioritisation of tasks.35 ,37 In addition, we did not specifically monitor whether participants attended postoperative rehabilitation.


The current study showed that people who underwent ACL reconstruction had a reduced ability to control movement of the body in the AP direction during a single leg balance task for both the operated and non-operated limbs. Implementing a secondary task increased CoP path length in the ACL-reconstructed and control groups by similar amounts, suggesting that the addition of a controlled perturbation to a balance task does not amplify the magnitude of balance impairment in this clinical population. Further research should examine balance tasks that place varying degrees of physical and cognitive demands on the participants to further elucidate the balance deficits present in people who have suffered an ACL injury.

What are the new findings

  • Postural control was evaluated using a novel secondary task to elucidate the strategies involved.

  • This study is one of few to match participants and controls for age, gender and physical activity level when evaluating postural control.

  • Anterior cruciate ligament (ACL)-reconstructed patients showed a reduced ability to control movement of the body particularly in the anterior–posterior direction during single-limb balance tasks.

  • The addition of a secondary task significantly increased centre of pressure path length in all test groups, although the magnitude of this increase was similar in both the ACL-reconstructed and control groups.

Impact on clinical practice

  • The Nintendo Wii Balance Board provided an objective method for quantifying defects in postural control following anterior cruciate ligament (ACL) reconstruction.

  • Devices such as this may represent an example of an inexpensive and portable tool which enables clinicians to more objectively measure postural control across a range of clinical settings.

  • Such testing may help identify patients with deficits in postural control who could benefit from additional targeted neuromuscular training to minimize the risk of further injury.


View Abstract


  • Contributors All authors made contributions to conception and design. BH and RC collected the data. BH, RC and KW analysed the data. BH drafted the article and all authors revised it critically for important intellectual content. All authors approved the version to be published.

  • Competing interests None.

  • Ethics approval Faculty of Health Sciences La Trobe University.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • ▸ References to this paper are available online at

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.