- Email: [email protected]
Association between seated postural control and gait speed in knee osteoarthritis
Gait & Posture 37 (2013) 413–418
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Association between seated postural control and gait speed in knee osteoarthritis Yong-Hao Pua a,*, Ross A. Clark b, Peck-Hoon Ong a, Adam L. Bryant b, Ngai-Nung Lo c, Zhiqi Liang a a
Department of Physiotherapy, Singapore General Hospital, Singapore Department of Physiotherapy, The University of Melbourne, Australia c Department of Orthopaedic Surgery, Singapore General Hospital, Singapore b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 9 August 2011 Received in revised form 26 April 2012 Accepted 21 August 2012
The purpose of this study was to evaluate, in patients with knee osteoarthritis, whether seated postural control is a multivariate predictor of gait speed, after adjusting for the effects of conventional knee impairments. Sixty-seven patients with knee osteoarthritis awaiting total knee replacement participated. To measure seated postural control, patients sat on a balance board, and the centre-ofpressure (COP) measures calculated in the anterior–posterior (AP) and medio-lateral (ML) directions were standard deviation (SD) and mean frequency (MF). Isometric knee extensor strength was measured using an isokinetic dynamometer; knee flexion range-of-motion, an extendable goniometer; and knee pain intensity, a numeric pain rating scale. Fast-pace gait speed was assessed by the 10-m walk test and a poor gait speed was defined at a cutoff value of 1.0 m/s. At the univariate level, the seated COP measures, with the exception of AP–MF, discriminated between patients with and without poor gait speed; however, only ML–MF retained its predictive value in multivariable analyses adjusted for demographic, anthropometric, and knee impairment measures. These findings suggest that seated postural control may be an important correlate of physical function in patients with knee OA and that greater emphasis in the assessment of trunk control may be warranted in this population. ß 2012 Elsevier B.V. All rights reserved.
Keywords: Biomechanics Knee Balance Mobility
1. Introduction What are the modifiable physical correlates of gait performance in older adults with symptomatic knee osteoarthritis (OA)? Knee OA, along with hip OA, adversely affects walking ability in older adults more than any other diseases [1], and a constellation of OArelated physical impairments affects walking speed, in which knee pain [2], muscle weakness [3] and loss of knee range-of-motion [4] have been implicated. Thus far, no knee OA studies have investigated the role of seated postural control – an index of active trunk control [5–7] – in influencing gait performance; yet, there are reasons to think that seated postural control may be an important correlate. First, from the biomechanical perspective, walking is a dynamic postural control task which requires the control of multiple joints such that the swing limb is placed under the falling centre-of-mass [8]. Since the trunk constitutes more than half the body mass [8], it seems reasonable to suggest that active trunk control is crucial in maintaining stability during walking [9].
* Corresponding author at: Department of Physiotherapy, Singapore General Hospital, Outram Road, Singapore 169608, Singapore. Tel.: +65 6321 4132. E-mail address: [email protected] (Y.-H. Pua). 0966-6362/$ – see front matter ß 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2012.08.014
Second, when considering the gait strategies of patients with knee OA, accumulating evidence suggests that these patients walked with increased lateral trunk lean towards the involved side and showed subtle increased range-of-motion in the pelvis [10,11]. Furthermore, Wang et al. [12] recently performed a crosssectional, comparative gait study on older adults with and without knee OA and observed that older adults with knee OA had greater centre-of-mass accelerations in the frontal plane which, in turn, resulted in more ‘‘jerky’’ frontal movements of the centre-of-mass. Altogether, these gait adaptations – ostensibly present to reduce knee joint loading – place greater demands on active trunk control to arrest the excessive lateral linear momentum of the trunk [12,13]. Or, put otherwise, a smooth gait progression depends on the patient’s ability to actively control rapid (jerky) trunk or centre-of-mass movements during walking. Despite these arguments, it remains to be demonstrated whether active trunk control and gait performance are empirically linked in knee OA. To the extent that the development of assessment tools and interventions to improve physical function may be aided by a better understanding of the association between trunk control and gait performance in knee OA, we conducted this cross-sectional study as a first step to explore the independent association between seated postural control and gait speed in a sample of patients with knee OA. Given the myriad of centre-of-pressure
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
414
(COP) measures that can be extracted from a seated postural control test [7], we focused on one time-domain measure (COP standard deviation) and one frequency-domain measure (COP mean frequency) to avoid spurious results. COP standard deviation is a conventional measure in most posturography studies [14] whilst more relevant to our focus, COP mean frequency was chosen as a potential measure of the ability of the body to produce rapid, corrective trunk movements [15]. These measures have been applied in older adults with and without neurological conditions [7,16] but we know of no studies of seated postural control in knee OA. 2. Methods 2.1. Patients The study sample comprised 67 patients with primary knee OA undergoing unilateral total knee replacement at a tertiary institution in Singapore from June 2010 to January 2011. Patients were recruited within a month before their surgery. Patients were excluded if they (i) were unable to walk 10 m independently without an assistive device, (ii) had significant back or other joint pain, or (iii) had any medical conditions that would compromise physical function or affect their abilities to complete testing. The SingHealth Institutional Review Board approved this study. Patients attended a test session at the outpatient physiotherapy department after providing written informed consent. Prior to the physical assessment, patients rated their knee pain using the visual numeric pain scale. Specifically, knee pain was rated on an 11-point scale, with 0 indicating ‘no pain’ and 10 indicating ‘worst pain ever experienced’. Two pain ratings – at rest and during activity over the past 24 h – were obtained and a composite average was calculated. The visual numeric pain scale has been shown to be a reliable and valid measure of pain in patients with OA [17]. 2.2. Physical measures Testing was performed in the following order: (i) knee flexion range-of-motion test, (ii) seated postural control test, (iii) gait speed test, and (iv) knee extensor strength test. Patients had their involved knees tested for the range-of-motion and strength tests. Patients were provided with rest periods as requested. A Gollehon extendable goniometer (Lafayette Instrument Company, Lafayette, IN) was used to measure knee flexion range-of-motion with the patient in the supine position. The axis of the goniometer was placed on the femoral lateral epicondyle. The proximal arm of the goniometer was placed directly on the greater trochanter of the femur; the distal arm of the goniometer was placed directly on the lateral malleolus of the ankle. Patients were asked to slide their heels towards the buttocks and maximum knee flexion range-of-motion was recorded. To measure seated postural control, we designed a test using the Wii Balance Board (WBB) (Nintendo, Kyoto, Japan). Specifically, the patient sat unsupported on the WBB which was placed on a treatment plinth (Fig. 1). The plinth height was
adjusted such that the patients’ hips and knees were flexed at 908 and their feet placed, shoulder-width apart, on a step-board. Patients were instructed to ‘‘look straight ahead, hold your arms across your chest and sit as quietly as possible.’’ Two 30-s trials were performed, with a rest period of 30 s between trials. The WBB was interfaced with a laptop computer using Labview (National Instruments, Austin, TX, USA), and it was calibrated using known loads placed at different positions on the WBB [18]. We have previously demonstrated that centreof-pressure (COP) measures from the WBB possessed good concurrent validity with those from a laboratory-grade forceplate [18]. To determine the frequency bandwidth which optimized the signal-to-noise ratio, the raw data for multiple trials were interpolated to 100 Hz and examined in the time–frequency domain using an analytic wavelet transform. Owing perhaps to the wide base-of-support afforded by the seated test, we observed little to no signal energy at frequencies exceeding 6 Hz [6]. Accordingly, we utilized a 6.25 Hz low pass filter that comprised a Coiflet-5 wavelet, and detail levels above this threshold frequency were subsequently removed. Of interest, wavelet-based filters were used because they are particularly effective at removing noise whilst maintaining signal integrity [19]. In this study, the COP standard deviation (SD) and mean frequency (MF) along both anteroposterior (AP) and mediolateral (ML) planes were used. The COP SD – a time-domain measure – was calculated as the variability of the COP around the mean position. The COP MF – a frequency-domain measure – was calculated as the average mean instantaneous frequency using a complex Morlet continuous wavelet transform. Specifically, the calculation of COP MF involved (i) obtaining the mean instantaneous frequency in 10 ms time scales for the entire trial, (ii) removing 5 s at both ends of the trial to remove the windowing effects, and (iii) computing the average frequency of the remaining time bins. Of interest, although the COP MF is generally analogous to the mean power frequency (MPF) derived from a fast Fourier transform, the use of interpolation and continuous wavelet transform overcomes the stationarity limitation (unlike the fast Fourier transform) and, importantly, has been shown to accurately reflect the signal frequency content [20]. Because the COP SD represents the variability of the COP in maintaining a mean quiet sitting position whilst the COP MF represents the average movement frequency underlying a person’s seated postural sway, a high COP MF and a low COP SD represent fast movements (i.e. high frequency) occurring in positions close to the mean COP position (i.e. a low standard deviation). This may be conventionally assumed to indicate that the person is able to utilize rapid, reflexive, and unconscious control processes to keep the COP within a small, defined boundary [21]. To measure gait speed, patients were timed using a stopwatch as they walked along a 10-m walkway at a fast pace. Patients stood directly behind the start line and were clocked from the time the first foot crossed the start line until the lead foot crossed the finish line. Patients were instructed to ‘‘walk as quickly as possible, but safely’’ and to finish at least 2 m past the finish line to eliminate the deceleration effects from stopping the walk. Each patient performed two valid trials and the better trial was taken for analysis. To measure knee extensor strength, maximal volitional isometric contraction of the knee extensors at 758 of knee flexion was obtained on a Biodex isokinetic dynamometer (Shirley, NY, USA). The patient was tested in a seated position with the hip at 908 of flexion, and the axis of rotation of the dynamometer lever arm was aligned to the femoral lateral condyle whilst the lever arm was secured to the tibia just proximal to the medial malleolus via an ankle cuff. Before testing the knee extensors, the gravity compensation procedure was performed by measuring the patient’s passive extremity weight at 308 of knee flexion. Following a warm-up comprising one submaximal and one maximal contraction, all patients performed two 5-s maximal trials with a 1-min rest interval. Patients were instructed to extend their knee as fast and forcefully as possible for at least 2–3 s. During testing, strong verbal encouragement was given. In each trial, the peak torque normalized to the patient’s body mass was recorded and the higher measurement of two valid trials was analysed.
2.3. Statistical analyses
Fig. 1. Assessment of seated postural control.
We used descriptive statistics to characterize the study sample: we used means with SDs for continuous variables and frequencies with percentages for categorical variables. We dichotomized gait speed at 1.0 m/s [22] and defined a patient as having poor gait speed when his or her fast-pace gait speed was below 1.0 m/s. Patients with versus without poor gait speed were compared on sociodemographic and functional characteristics using Student’s t-test or chi-square test, as appropriate. Because the seated COP data were not normally distributed, we assessed between-group differences using the exact Wilcoxon–Mann–Whitney test. To examine the multivariable associations of gait performance with the seated COP ML and AP measures, we used separate multivariable regression models and included in each model covariables that we believed a priori to have an association with gait speed – namely, sex, age, body weight, knee pain, knee extensor strength, and knee flexion range of motion. All independent variables were entered simultaneously. To facilitate the interpretation of effect sizes, we calculated squared semi-partial correlation coefficients which represent the unique proportion of the total variance in the dependent variable accounted for by a given independent variable.
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418 All analyses were done in R software, version 2.12.1 (R Foundation, Vienna, Austria). Statistical significance was determined at the 2-sided 0.05 level.
3. Results Table 1 summarizes the participants’ sociodemographic and functional characteristics. Of the total sample, 28 (42%) patients met the criterion for poor gait speed. Patients with poor gait speed were similar to patients without poor gait speed in levels of knee pain, but they were older and comprised mainly women. They also had lower levels of knee extensor strength and knee flexion range of motion. In addition, the seated COP measures – with the exception of AP–MF – were significantly different between the two groups. As shown in Fig. 2, ML–MF was significantly higher (6.4%) in patients without poor gait speed whilst ML–SD and AP–SD were both significantly lower (14.8% and 29.8%, respectively). Table 2 shows the multivariable models of the association of fast-pace gait speed with the seated ML and AP COP measures. In
415
both models, age and gender were consistently associated with gait speed, whilst knee extensor strength and knee flexion range of motion were significantly, or nearly significantly (p = 0.10) associated with gait speed. Adjusting for demographic, anthropometric, and knee impairment measures, ML–MF – but not ML–SD or the AP measures – emerged as a multivariable predictor (p < 0.01). Supporting this, partial correlation analyses (Fig. 3) revealed that amongst the various seated COP measures, ML–MF showed the strongest association with gait speed and it explained comparable (or more) variance in gait speed as did the knee impairment measures. 4. Discussion In a sample of adults with end-stage knee OA, we explored the relationship between seated postural control and gait speed. We found that although the seated COP measures, with the exception of AP–MF, discriminated between patients with and without poor
Fig. 2. Notched box-and-whisker plots showing 25th, 50th, and 75th percentiles of seated COP measures by gait-speed groups. Individual COP values are represented by closed circles overlaying the plots. Open circles represent outliers. *Significant difference, as assessed using the exact Wilcoxon–Mann–Whitney test. (A) ML–SD; (B) ML–MF; (C) AP–SD and (D) AP–MF.
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
416
Table 1 Sociodemographic and functional characteristics of the study group. Variables Age (years) No. of women (%) Height (m) Mass (kg) BMI (kg/m2) No. of bilateral symptoms (%) Gait speed (m/s) Knee impairments Knee extensor strength (Nm/kg) Knee pain (/10) Knee flexion range of motion (8) Seated COP measuresb ML–SD (cm) ML–MF (Hz) AP–SD (cm) AP–MF (Hz)
All patients (n = 67)
Gait speed 1.0 m/s (n = 39)
Gait speed < 1.0 m/s (n = 28)
p-Valuea
67.19 7.5 46 (69) 1.57 0.1 69.7 11.2 28.5 4.6 44 (66) 1.03 0.29
64.1 6.6 21 (54) 1.59 0.1 69.4 12.2 27.5 4.4 25 (64) 1.23 0.16
71.5 6.3 25 (89) 1.54 0.1 69.7 11.2 29.8 4.6 19 (68) 0.75 0.17
<0.001 0.002 0.001 0.81 0.04 0.75 <0.001
1.13 0.45 3.8 1.7 114 13
1.25 0.46 3.8 1.7 118 10
1.03 0.29 3.7 1.8 108 15
0.01 0.74 0.003
0.105 0.045 1.30 0.14 0.086 0.051 1.45 0.15
0.098 0.04 1.33 0.12 0.073 0.04 1.47 0.14
0.115 0.04 1.25 0.14 0.104 0.05 1.43 0.15
0.04 0.02 0.01 0.10
Values are mean SD unless otherwise indicated. BMI: body mass index; ML: mediolateral; AP: anteroposterior; MF: mean frequency; SD: standard deviation; COP: centre of pressure. a Student’s t-test was used for comparison of continuous variables. x2 test was used for comparison of categorical variables. b Between group differences in seated COP measures were assessed using exact Wilcoxon–Mann–Whitney test.
Fig. 3. Squared semi-partial correlation coefficients for the explanatory variables predicting fastpace gait speed. (A) Multivariable model involving the seated ML measures; (B) multivariable model involving the seated AP measures. Table 2 Multivariable models of the association of fast-paced gait speed with seated COP measures. Variables Model 1 (adjusted R2 = 0.62y) Age Gender Body mass Knee pain Knee flexion range of motion Knee extensor strength ML–SD ML–MF Model 2 (adjusted R2 = 0.55y) Age Gender Body mass Knee pain Knee flexion range of motion Knee extensor strength AP–SD AP–MF
b SE
p-Value
0.015 0.004 0.25 0.06 0.0001 0.002 0.015 0.014 0.004 0.002 0.11 0.06 0.57 0.57 0.49 0.18
<0.0001 <0.0001 0.96 0.27 0.03 0.10 0.33 <0.01
0.017 0.004 0.23 0.06 0.0001 0.002 0.011 0.015 0.004 0.002 0.11 0.07 0.63 0.52 0.066 0.18
<0.0001 <0.001 0.97 0.46 0.06 0.10 0.23 0.71
Data are unstandardized regression coefficients standard errors. Gender (1 = men, 2 = women). ML: mediolateral; AP: anteroposterior; MF: mean frequency; SD: standard deviation. y p < 0.001.
gait speed at the univariate level, the significant associations with gait speed persisted only for ML–MF in the multivariable analyses. To our knowledge, the associations between seated postural control and gait performance have not been evaluated before in patients with knee OA, and certainly not in a multivariable manner. Our regression findings raise two questions. First, how do we explain the positive significant association between ML–MF and gait performance? Although we found no knee OA studies sufficiently similar to ours to make direct comparisons, one recent study of seated postural control in 331 older adults noted that greater overall MPF was (non-significantly) associated with decreasing odds of standing balance loss [7]. That study, however, did not isolate and examine ML–MPF in the regression analyses. On the other hand, using an unstable sitting protocol, van Dieen et al. [16] showed that ML–MPF – but not AP–MPF – increased monotonically from (i) older adults with Parkinson’s disease and a fall history to (ii) older adults with Parkinson’s disease and no fall history and to (i) healthy older adults. Other studies examined standing COP MPF measures in relation to fall risk in older adults and reported that higher levels of overall MPF [23] or ML–MPF [24] were associated with lower fall risk. Ostensibly, a high COP MF (with a concomitant low SD) reflects the use of rapid and tight
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
control processes, and the effects may be interpreted in terms of high frequency movement adjustments that enhance one’s standing postural control [15]. In the context of our study, we suggest that perhaps patients who had high COP MF values during the seated balance test were those with greater trunk proprioceptive acuity which, in turn, allowed these patients to make tighter, higher frequency trunk adjustments during walking to stabilize the trunk. Indirectly supporting this, van Dieen et al. [25] performed a cross-sectional, comparative study of seated balance performance in persons with and without low back pain. The authors observed lower frequency sway in persons with low back pain – a clinical group with ostensibly impaired trunk proprioceptive acuity and reduced trunk muscle performance. Granted, we acknowledge that we do not have data on trunk motion and trunk muscle activity of our patients and that drawing direct parallels between standing and seated balance performance is difficult. Additional studies are clearly needed to verify our results and to elucidate the mechanisms linking a high seated ML–MF with faster gait speed. The second question is that in contrast to the significant findings for ML–MF, how do we explain the null findings for the AP measures? Although we again found no comparable studies, two indirect lines of evidence may explain our findings. First, robotic gait models [26] and human gait analyses [26–28] have shown that during walking, passive dynamic stability exists in the sagittal (AP) but not the frontal (ML) plane. Indeed, humans showed greater step width variability than step length variability during walking [26–28], and given also that step width variability was associated with trunk motion variability [29], integrative trunk control may be particularly important in the frontal plane. Further supporting this notion, in two recent carefully controlled biomechanical studies on 12 healthy adults [27,28], McAndrews and colleagues found that the trunk movements of these subjects during treadmill walking were more unstable to both visual and mechanical perturbations applied in the ML direction than to those applied in the AP direction. Collectively, our results and those of other studies suggest that active trunk control – particularly in the frontal plane – wields an important influence on gait performance. Second, and more specifically, the compensatory gait strategies of patients with knee OA may offer yet another explanation for our findings. Specifically, these gait strategies include, inter alia, contralateral cane use, increased step width, and increased lateral trunk lean [10,11]. Inasmuch as these gait strategies involved increased COM acceleration and repetitive weight-shifts in the frontal plane [12], it follows that lateral trunk control is a crucial correlate of gait performance. That said, because we did not formally document the gait strategies in our sample, our inferences are at best speculative and future studies should examine whether seated postural control interacted with the different gait strategies to influence gait performance. Our findings have implications. First, perhaps our most compelling finding is that seated postural control was an independent correlate of gait performance, after adjusting for the effects of conventional knee impairments. Hitherto, knee OA studies examining pelvic control have narrowly focused on the lateral hip muscles [30,31], and very little is said about the close functional associations between the trunk and the pelvis. Our findings, if replicated, indicate that seated postural control should be considered in the clinical assessment of patients with knee OA. Furthermore, addressing possible trunk control impairments – for example, improving trunk musculature performance [32] – as part of the overall OA management could potentially enhance physical function. Second, from the research viewpoint, our results are methodologically important. Specifically, by demonstrating the convergent validity of the seated COP measures with gait performance, our results support the repurposing of the Wii
417
Board – a cheaply available and portable gaming device – as a practicable and novel assessment tool of seated postural control. Indeed, the current paucity of knee OA research on trunk control is probably, in part, a reflection of the logistical and technical difficulties of assessing seated postural control. For this reason, if our results were replicated, we envisage the potential application of the Wii Board in both clinical and research protocols. Our study has limitations in addition to those already stated above. First, our results may be affected by reverse causation due to their cross-sectional nature and intervention studies are clearly needed to establish the presence and direction of causation. Second, we studied patients with end-stage knee OA and it is uncertain if our findings are invariantly applicable to patients with mild to moderate knee OA. Third, we measured only gait performance, and it would be informative to evaluate other functional tasks – for example, stair climbing and sit-to-stand – given their potential abilities to tax trunk control in multiple planes. Relatedly, it would also be desirable to evaluate trunk control using an unstable sitting protocol given its ability to provide greater postural challenge to the trunk [7,16]. 5. Conclusions In patients with symptomatic knee OA, seated postural control – particularly in the ML direction – was associated with gait speed performance. Although our findings are strictly correlational, they suggest that seated postural control is an important correlate of physical function and that greater emphasis in the assessment of trunk control may be warranted. Finally, intervention studies are necessary to determine whether improvements in trunk control could enhance gait performance in patients with knee OA. Acknowledgements We thank our patients for donating their time. We also thank our colleagues, Bee-Yee Tan, Hanniel Lim, Felicia Seet, Jennifer Liaw, Hwei-Chi Chong, and William Yeo, from the Singapore General Hospital, for supporting this study. Finally, we thank the orthopaedic surgeons from the Singapore General Hospital for allowing us access to their patients. Conflict of interest None. References [1] Guccione AA, Felson DT, Anderson JJ, Anthony JM, Zhang YQ, Wilson PWF, et al. The effects of specific medical conditions on the functional limitations of elders in the Framingham Study. American Journal of Public Health 1994;84:351–8. [2] Sowers M, Jannausch ML, Gross M, Karvonen-Gutierrez CA, Palmieri RM, Crutchfield M, et al. Performance-based physical functioning in African-American and Caucasian women at midlife: considering body composition, quadriceps strength, and knee osteoarthritis. American Journal of Epidemiology 2006;163:950–8. [3] Schmitt LC, Rudolph KS. Influences on knee movement strategies during walking in persons with medial knee osteoarthritis. Arthritis and Rheumatism 2007;57:1018–26. [4] Maly MR, Costigan PA, Olney SJ. Role of knee kinematics and kinetics on performance and disability in people with medial compartment knee osteoarthritis. Clinical Biomechanics (Bristol Avon) 2006;21:1051–9. [5] Cholewicki J, Polzhofer GK, Radebold A. Postural control of trunk during unstable sitting. Journal of Biomechanics 2000;33:1733–7. [6] Bennett BC, Abel MF, Granata KP. Seated postural control in adolescents with idiopathic scoliosis. Spine (Philadelphia PA 1976) 2004;29:E449–54. [7] van Dieen JH, Koppes LL, Twisk JW. Postural sway parameters in seated balancing; their reliability and relationship with balancing performance. Gait and Posture 2010;31:42–6. [8] Winter DA. Biomechanics and motor control of human movement, 4th ed., New York: John Wiley & Sons Inc.; 2009. [9] Kang HG, Dingwell JB. Dynamic stability of superior vs. inferior segments during walking in young and older adults. Gait and Posture 2009;30:260–3.
418
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
[10] Hunt MA, Wrigley TV, Hinman RS, Bennell KL. Individuals with severe knee osteoarthritis (OA) exhibit altered proximal walking mechanics compared with individuals with less severe OA and those without knee pain. Arthritis Care and Research (Hoboken) 2010;62:1426–32. [11] Mundermann A, Asay JL, Mundermann L, Andriacchi TP. Implications of increased medio-lateral trunk sway for ambulatory mechanics. Journal of Biomechanics 2008;41:165–70. [12] Wang TM, Hsu WC, Chang CF, Hu CC, Lu TW. Effects of knee osteoarthritis on body’s center of mass motion in older adults during level walking. Biomedical Engineering Applications Basis and Communications 2010; 22:205–12. [13] Kaya BK, Krebs DE, Riley PO. Dynamic stability in elders: momentum control in locomotor ADL. Journals of Gerontology Series A Biological Sciences and Medical Sciences 1998;53:M126–34. [14] Visser JE, Carpenter MG, van der Kooij H, Bloem BR. The clinical utility of posturography. Clinical Neurophysiology 2008;119:2424–36. [15] Adkin AL, Frank JS, Carpenter MG, Peysar GW. Postural control is scaled to level of postural threat. Gait and Posture 2000;12:87–93. [16] van der Burg JC, van Wegen EE, Rietberg MB, Kwakkel G, van Dieen JH. Postural control of the trunk during unstable sitting in Parkinson’s disease. Parkinsonism & Related Disorders 2006;12:492–8. [17] Ritter PL, Gonzalez VM, Laurent DD, Lorig KR. Measurement of pain using the visual numeric scale. Journal of Rheumatology 2006;33:574–80. [18] Clark RA, Bryant AL, Pua YH, McCrory P, Bennell K, Hunt M. Validity and reliability of the Nintendo Wii Balance Board for assessment of standing balance. Gait and Posture 2010;31:307–10. [19] Ismail AR, Asfour SS. Discrete wavelet transform: a tool in smoothing kinematic data. Journal of Biomechanics 1999;32:317–21. [20] Thiebaut C, Roques S. Time-scale and time-frequency analyses of irregularly sampled astronomical time series. EURASIP Journal on Advances in Signal Processing 2005;15:2486–99. [21] Chiviacowsky S, Wulf G, Wally R. An external focus of attention enhances balance learning in older adults. Gait and Posture 2010;32:572–5.
[22] Brown M, Sinacore DR, Binder EF, Kohrt WM. Physical and performance measures for the identification of mild to moderate frailty. Journals of Gerontology Series A Biological Sciences and Medical Sciences 2000;55:M350–5. [23] Bigelow KE, Berme N. Development of a protocol for improving the clinical utility of posturography as a fall-risk screening tool. Journals of Gerontology Series A Biological Sciences and Medical Sciences 2011;66:228. [24] Williams HG, McClenaghan BA, Dickerson J. Spectral characteristics of postural control in elderly individuals. Archives of Physical Medicine and Rehabilitation 1997;78:737–44. [25] van Dieen JH, Koppes LL, Twisk JW. Low back pain history and postural sway in unstable sitting. Spine (Philadelphia PA 1976) 2010;35:812–7. [26] Bauby CE, Kuo AD. Active control of lateral balance in human walking. Journal of Biomechanics 2000;33:1433–40. [27] McAndrew PM, Dingwell JB, Wilken JM. Walking variability during continuous pseudo-random oscillations of the support surface and visual field. Journal of Biomechanics 2010;43:1470–5. [28] McAndrew PM, Wilken JM, Dingwell JB. Dynamic stability of human walking in visually and mechanically destabilizing environments. Journal of Biomechanics 2011;44:644–9. [29] Hurt CP, Rosenblatt N, Crenshaw JR, Grabiner MD. Variation in trunk kinematics influences variation in step width during treadmill walking by older and younger adults. Gait and Posture 2010;31:461–4. [30] Sled EA, Khoja L, Deluzio KJ, Olney SJ, Culham EG. Effect of a home program of hip abductor exercises on knee joint loading, strength, function, and pain in people with knee osteoarthritis: a clinical trial. Physical Therapy 2010;90:895–904. [31] Bennell KL, Hunt MA, Wrigley TV, Hunter DJ, McManus FJ, Hodges PW, et al. Hip strengthening reduces symptoms but not knee load in people with medial knee osteoarthritis and varus malalignment: a randomised controlled trial. Osteoarthritis and Cartilage 2010;18:621–8. [32] Suri P, Kiely DK, Leveille SG, Frontera WR, Bean JF. Increased trunk extension endurance Is associated with meaningful improvement in balance among older adults with mobility problems. Archives of Physical Medicine and Rehabilitation 2011;92:1038–43.
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Association between seated postural control and gait speed in knee osteoarthritis Yong-Hao Pua a,*, Ross A. Clark b, Peck-Hoon Ong a, Adam L. Bryant b, Ngai-Nung Lo c, Zhiqi Liang a a
Department of Physiotherapy, Singapore General Hospital, Singapore Department of Physiotherapy, The University of Melbourne, Australia c Department of Orthopaedic Surgery, Singapore General Hospital, Singapore b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 9 August 2011 Received in revised form 26 April 2012 Accepted 21 August 2012
The purpose of this study was to evaluate, in patients with knee osteoarthritis, whether seated postural control is a multivariate predictor of gait speed, after adjusting for the effects of conventional knee impairments. Sixty-seven patients with knee osteoarthritis awaiting total knee replacement participated. To measure seated postural control, patients sat on a balance board, and the centre-ofpressure (COP) measures calculated in the anterior–posterior (AP) and medio-lateral (ML) directions were standard deviation (SD) and mean frequency (MF). Isometric knee extensor strength was measured using an isokinetic dynamometer; knee flexion range-of-motion, an extendable goniometer; and knee pain intensity, a numeric pain rating scale. Fast-pace gait speed was assessed by the 10-m walk test and a poor gait speed was defined at a cutoff value of 1.0 m/s. At the univariate level, the seated COP measures, with the exception of AP–MF, discriminated between patients with and without poor gait speed; however, only ML–MF retained its predictive value in multivariable analyses adjusted for demographic, anthropometric, and knee impairment measures. These findings suggest that seated postural control may be an important correlate of physical function in patients with knee OA and that greater emphasis in the assessment of trunk control may be warranted in this population. ß 2012 Elsevier B.V. All rights reserved.
Keywords: Biomechanics Knee Balance Mobility
1. Introduction What are the modifiable physical correlates of gait performance in older adults with symptomatic knee osteoarthritis (OA)? Knee OA, along with hip OA, adversely affects walking ability in older adults more than any other diseases [1], and a constellation of OArelated physical impairments affects walking speed, in which knee pain [2], muscle weakness [3] and loss of knee range-of-motion [4] have been implicated. Thus far, no knee OA studies have investigated the role of seated postural control – an index of active trunk control [5–7] – in influencing gait performance; yet, there are reasons to think that seated postural control may be an important correlate. First, from the biomechanical perspective, walking is a dynamic postural control task which requires the control of multiple joints such that the swing limb is placed under the falling centre-of-mass [8]. Since the trunk constitutes more than half the body mass [8], it seems reasonable to suggest that active trunk control is crucial in maintaining stability during walking [9].
* Corresponding author at: Department of Physiotherapy, Singapore General Hospital, Outram Road, Singapore 169608, Singapore. Tel.: +65 6321 4132. E-mail address: [email protected] (Y.-H. Pua). 0966-6362/$ – see front matter ß 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2012.08.014
Second, when considering the gait strategies of patients with knee OA, accumulating evidence suggests that these patients walked with increased lateral trunk lean towards the involved side and showed subtle increased range-of-motion in the pelvis [10,11]. Furthermore, Wang et al. [12] recently performed a crosssectional, comparative gait study on older adults with and without knee OA and observed that older adults with knee OA had greater centre-of-mass accelerations in the frontal plane which, in turn, resulted in more ‘‘jerky’’ frontal movements of the centre-of-mass. Altogether, these gait adaptations – ostensibly present to reduce knee joint loading – place greater demands on active trunk control to arrest the excessive lateral linear momentum of the trunk [12,13]. Or, put otherwise, a smooth gait progression depends on the patient’s ability to actively control rapid (jerky) trunk or centre-of-mass movements during walking. Despite these arguments, it remains to be demonstrated whether active trunk control and gait performance are empirically linked in knee OA. To the extent that the development of assessment tools and interventions to improve physical function may be aided by a better understanding of the association between trunk control and gait performance in knee OA, we conducted this cross-sectional study as a first step to explore the independent association between seated postural control and gait speed in a sample of patients with knee OA. Given the myriad of centre-of-pressure
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
414
(COP) measures that can be extracted from a seated postural control test [7], we focused on one time-domain measure (COP standard deviation) and one frequency-domain measure (COP mean frequency) to avoid spurious results. COP standard deviation is a conventional measure in most posturography studies [14] whilst more relevant to our focus, COP mean frequency was chosen as a potential measure of the ability of the body to produce rapid, corrective trunk movements [15]. These measures have been applied in older adults with and without neurological conditions [7,16] but we know of no studies of seated postural control in knee OA. 2. Methods 2.1. Patients The study sample comprised 67 patients with primary knee OA undergoing unilateral total knee replacement at a tertiary institution in Singapore from June 2010 to January 2011. Patients were recruited within a month before their surgery. Patients were excluded if they (i) were unable to walk 10 m independently without an assistive device, (ii) had significant back or other joint pain, or (iii) had any medical conditions that would compromise physical function or affect their abilities to complete testing. The SingHealth Institutional Review Board approved this study. Patients attended a test session at the outpatient physiotherapy department after providing written informed consent. Prior to the physical assessment, patients rated their knee pain using the visual numeric pain scale. Specifically, knee pain was rated on an 11-point scale, with 0 indicating ‘no pain’ and 10 indicating ‘worst pain ever experienced’. Two pain ratings – at rest and during activity over the past 24 h – were obtained and a composite average was calculated. The visual numeric pain scale has been shown to be a reliable and valid measure of pain in patients with OA [17]. 2.2. Physical measures Testing was performed in the following order: (i) knee flexion range-of-motion test, (ii) seated postural control test, (iii) gait speed test, and (iv) knee extensor strength test. Patients had their involved knees tested for the range-of-motion and strength tests. Patients were provided with rest periods as requested. A Gollehon extendable goniometer (Lafayette Instrument Company, Lafayette, IN) was used to measure knee flexion range-of-motion with the patient in the supine position. The axis of the goniometer was placed on the femoral lateral epicondyle. The proximal arm of the goniometer was placed directly on the greater trochanter of the femur; the distal arm of the goniometer was placed directly on the lateral malleolus of the ankle. Patients were asked to slide their heels towards the buttocks and maximum knee flexion range-of-motion was recorded. To measure seated postural control, we designed a test using the Wii Balance Board (WBB) (Nintendo, Kyoto, Japan). Specifically, the patient sat unsupported on the WBB which was placed on a treatment plinth (Fig. 1). The plinth height was
adjusted such that the patients’ hips and knees were flexed at 908 and their feet placed, shoulder-width apart, on a step-board. Patients were instructed to ‘‘look straight ahead, hold your arms across your chest and sit as quietly as possible.’’ Two 30-s trials were performed, with a rest period of 30 s between trials. The WBB was interfaced with a laptop computer using Labview (National Instruments, Austin, TX, USA), and it was calibrated using known loads placed at different positions on the WBB [18]. We have previously demonstrated that centreof-pressure (COP) measures from the WBB possessed good concurrent validity with those from a laboratory-grade forceplate [18]. To determine the frequency bandwidth which optimized the signal-to-noise ratio, the raw data for multiple trials were interpolated to 100 Hz and examined in the time–frequency domain using an analytic wavelet transform. Owing perhaps to the wide base-of-support afforded by the seated test, we observed little to no signal energy at frequencies exceeding 6 Hz [6]. Accordingly, we utilized a 6.25 Hz low pass filter that comprised a Coiflet-5 wavelet, and detail levels above this threshold frequency were subsequently removed. Of interest, wavelet-based filters were used because they are particularly effective at removing noise whilst maintaining signal integrity [19]. In this study, the COP standard deviation (SD) and mean frequency (MF) along both anteroposterior (AP) and mediolateral (ML) planes were used. The COP SD – a time-domain measure – was calculated as the variability of the COP around the mean position. The COP MF – a frequency-domain measure – was calculated as the average mean instantaneous frequency using a complex Morlet continuous wavelet transform. Specifically, the calculation of COP MF involved (i) obtaining the mean instantaneous frequency in 10 ms time scales for the entire trial, (ii) removing 5 s at both ends of the trial to remove the windowing effects, and (iii) computing the average frequency of the remaining time bins. Of interest, although the COP MF is generally analogous to the mean power frequency (MPF) derived from a fast Fourier transform, the use of interpolation and continuous wavelet transform overcomes the stationarity limitation (unlike the fast Fourier transform) and, importantly, has been shown to accurately reflect the signal frequency content [20]. Because the COP SD represents the variability of the COP in maintaining a mean quiet sitting position whilst the COP MF represents the average movement frequency underlying a person’s seated postural sway, a high COP MF and a low COP SD represent fast movements (i.e. high frequency) occurring in positions close to the mean COP position (i.e. a low standard deviation). This may be conventionally assumed to indicate that the person is able to utilize rapid, reflexive, and unconscious control processes to keep the COP within a small, defined boundary [21]. To measure gait speed, patients were timed using a stopwatch as they walked along a 10-m walkway at a fast pace. Patients stood directly behind the start line and were clocked from the time the first foot crossed the start line until the lead foot crossed the finish line. Patients were instructed to ‘‘walk as quickly as possible, but safely’’ and to finish at least 2 m past the finish line to eliminate the deceleration effects from stopping the walk. Each patient performed two valid trials and the better trial was taken for analysis. To measure knee extensor strength, maximal volitional isometric contraction of the knee extensors at 758 of knee flexion was obtained on a Biodex isokinetic dynamometer (Shirley, NY, USA). The patient was tested in a seated position with the hip at 908 of flexion, and the axis of rotation of the dynamometer lever arm was aligned to the femoral lateral condyle whilst the lever arm was secured to the tibia just proximal to the medial malleolus via an ankle cuff. Before testing the knee extensors, the gravity compensation procedure was performed by measuring the patient’s passive extremity weight at 308 of knee flexion. Following a warm-up comprising one submaximal and one maximal contraction, all patients performed two 5-s maximal trials with a 1-min rest interval. Patients were instructed to extend their knee as fast and forcefully as possible for at least 2–3 s. During testing, strong verbal encouragement was given. In each trial, the peak torque normalized to the patient’s body mass was recorded and the higher measurement of two valid trials was analysed.
2.3. Statistical analyses
Fig. 1. Assessment of seated postural control.
We used descriptive statistics to characterize the study sample: we used means with SDs for continuous variables and frequencies with percentages for categorical variables. We dichotomized gait speed at 1.0 m/s [22] and defined a patient as having poor gait speed when his or her fast-pace gait speed was below 1.0 m/s. Patients with versus without poor gait speed were compared on sociodemographic and functional characteristics using Student’s t-test or chi-square test, as appropriate. Because the seated COP data were not normally distributed, we assessed between-group differences using the exact Wilcoxon–Mann–Whitney test. To examine the multivariable associations of gait performance with the seated COP ML and AP measures, we used separate multivariable regression models and included in each model covariables that we believed a priori to have an association with gait speed – namely, sex, age, body weight, knee pain, knee extensor strength, and knee flexion range of motion. All independent variables were entered simultaneously. To facilitate the interpretation of effect sizes, we calculated squared semi-partial correlation coefficients which represent the unique proportion of the total variance in the dependent variable accounted for by a given independent variable.
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418 All analyses were done in R software, version 2.12.1 (R Foundation, Vienna, Austria). Statistical significance was determined at the 2-sided 0.05 level.
3. Results Table 1 summarizes the participants’ sociodemographic and functional characteristics. Of the total sample, 28 (42%) patients met the criterion for poor gait speed. Patients with poor gait speed were similar to patients without poor gait speed in levels of knee pain, but they were older and comprised mainly women. They also had lower levels of knee extensor strength and knee flexion range of motion. In addition, the seated COP measures – with the exception of AP–MF – were significantly different between the two groups. As shown in Fig. 2, ML–MF was significantly higher (6.4%) in patients without poor gait speed whilst ML–SD and AP–SD were both significantly lower (14.8% and 29.8%, respectively). Table 2 shows the multivariable models of the association of fast-pace gait speed with the seated ML and AP COP measures. In
415
both models, age and gender were consistently associated with gait speed, whilst knee extensor strength and knee flexion range of motion were significantly, or nearly significantly (p = 0.10) associated with gait speed. Adjusting for demographic, anthropometric, and knee impairment measures, ML–MF – but not ML–SD or the AP measures – emerged as a multivariable predictor (p < 0.01). Supporting this, partial correlation analyses (Fig. 3) revealed that amongst the various seated COP measures, ML–MF showed the strongest association with gait speed and it explained comparable (or more) variance in gait speed as did the knee impairment measures. 4. Discussion In a sample of adults with end-stage knee OA, we explored the relationship between seated postural control and gait speed. We found that although the seated COP measures, with the exception of AP–MF, discriminated between patients with and without poor
Fig. 2. Notched box-and-whisker plots showing 25th, 50th, and 75th percentiles of seated COP measures by gait-speed groups. Individual COP values are represented by closed circles overlaying the plots. Open circles represent outliers. *Significant difference, as assessed using the exact Wilcoxon–Mann–Whitney test. (A) ML–SD; (B) ML–MF; (C) AP–SD and (D) AP–MF.
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
416
Table 1 Sociodemographic and functional characteristics of the study group. Variables Age (years) No. of women (%) Height (m) Mass (kg) BMI (kg/m2) No. of bilateral symptoms (%) Gait speed (m/s) Knee impairments Knee extensor strength (Nm/kg) Knee pain (/10) Knee flexion range of motion (8) Seated COP measuresb ML–SD (cm) ML–MF (Hz) AP–SD (cm) AP–MF (Hz)
All patients (n = 67)
Gait speed 1.0 m/s (n = 39)
Gait speed < 1.0 m/s (n = 28)
p-Valuea
67.19 7.5 46 (69) 1.57 0.1 69.7 11.2 28.5 4.6 44 (66) 1.03 0.29
64.1 6.6 21 (54) 1.59 0.1 69.4 12.2 27.5 4.4 25 (64) 1.23 0.16
71.5 6.3 25 (89) 1.54 0.1 69.7 11.2 29.8 4.6 19 (68) 0.75 0.17
<0.001 0.002 0.001 0.81 0.04 0.75 <0.001
1.13 0.45 3.8 1.7 114 13
1.25 0.46 3.8 1.7 118 10
1.03 0.29 3.7 1.8 108 15
0.01 0.74 0.003
0.105 0.045 1.30 0.14 0.086 0.051 1.45 0.15
0.098 0.04 1.33 0.12 0.073 0.04 1.47 0.14
0.115 0.04 1.25 0.14 0.104 0.05 1.43 0.15
0.04 0.02 0.01 0.10
Values are mean SD unless otherwise indicated. BMI: body mass index; ML: mediolateral; AP: anteroposterior; MF: mean frequency; SD: standard deviation; COP: centre of pressure. a Student’s t-test was used for comparison of continuous variables. x2 test was used for comparison of categorical variables. b Between group differences in seated COP measures were assessed using exact Wilcoxon–Mann–Whitney test.
Fig. 3. Squared semi-partial correlation coefficients for the explanatory variables predicting fastpace gait speed. (A) Multivariable model involving the seated ML measures; (B) multivariable model involving the seated AP measures. Table 2 Multivariable models of the association of fast-paced gait speed with seated COP measures. Variables Model 1 (adjusted R2 = 0.62y) Age Gender Body mass Knee pain Knee flexion range of motion Knee extensor strength ML–SD ML–MF Model 2 (adjusted R2 = 0.55y) Age Gender Body mass Knee pain Knee flexion range of motion Knee extensor strength AP–SD AP–MF
b SE
p-Value
0.015 0.004 0.25 0.06 0.0001 0.002 0.015 0.014 0.004 0.002 0.11 0.06 0.57 0.57 0.49 0.18
<0.0001 <0.0001 0.96 0.27 0.03 0.10 0.33 <0.01
0.017 0.004 0.23 0.06 0.0001 0.002 0.011 0.015 0.004 0.002 0.11 0.07 0.63 0.52 0.066 0.18
<0.0001 <0.001 0.97 0.46 0.06 0.10 0.23 0.71
Data are unstandardized regression coefficients standard errors. Gender (1 = men, 2 = women). ML: mediolateral; AP: anteroposterior; MF: mean frequency; SD: standard deviation. y p < 0.001.
gait speed at the univariate level, the significant associations with gait speed persisted only for ML–MF in the multivariable analyses. To our knowledge, the associations between seated postural control and gait performance have not been evaluated before in patients with knee OA, and certainly not in a multivariable manner. Our regression findings raise two questions. First, how do we explain the positive significant association between ML–MF and gait performance? Although we found no knee OA studies sufficiently similar to ours to make direct comparisons, one recent study of seated postural control in 331 older adults noted that greater overall MPF was (non-significantly) associated with decreasing odds of standing balance loss [7]. That study, however, did not isolate and examine ML–MPF in the regression analyses. On the other hand, using an unstable sitting protocol, van Dieen et al. [16] showed that ML–MPF – but not AP–MPF – increased monotonically from (i) older adults with Parkinson’s disease and a fall history to (ii) older adults with Parkinson’s disease and no fall history and to (i) healthy older adults. Other studies examined standing COP MPF measures in relation to fall risk in older adults and reported that higher levels of overall MPF [23] or ML–MPF [24] were associated with lower fall risk. Ostensibly, a high COP MF (with a concomitant low SD) reflects the use of rapid and tight
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
control processes, and the effects may be interpreted in terms of high frequency movement adjustments that enhance one’s standing postural control [15]. In the context of our study, we suggest that perhaps patients who had high COP MF values during the seated balance test were those with greater trunk proprioceptive acuity which, in turn, allowed these patients to make tighter, higher frequency trunk adjustments during walking to stabilize the trunk. Indirectly supporting this, van Dieen et al. [25] performed a cross-sectional, comparative study of seated balance performance in persons with and without low back pain. The authors observed lower frequency sway in persons with low back pain – a clinical group with ostensibly impaired trunk proprioceptive acuity and reduced trunk muscle performance. Granted, we acknowledge that we do not have data on trunk motion and trunk muscle activity of our patients and that drawing direct parallels between standing and seated balance performance is difficult. Additional studies are clearly needed to verify our results and to elucidate the mechanisms linking a high seated ML–MF with faster gait speed. The second question is that in contrast to the significant findings for ML–MF, how do we explain the null findings for the AP measures? Although we again found no comparable studies, two indirect lines of evidence may explain our findings. First, robotic gait models [26] and human gait analyses [26–28] have shown that during walking, passive dynamic stability exists in the sagittal (AP) but not the frontal (ML) plane. Indeed, humans showed greater step width variability than step length variability during walking [26–28], and given also that step width variability was associated with trunk motion variability [29], integrative trunk control may be particularly important in the frontal plane. Further supporting this notion, in two recent carefully controlled biomechanical studies on 12 healthy adults [27,28], McAndrews and colleagues found that the trunk movements of these subjects during treadmill walking were more unstable to both visual and mechanical perturbations applied in the ML direction than to those applied in the AP direction. Collectively, our results and those of other studies suggest that active trunk control – particularly in the frontal plane – wields an important influence on gait performance. Second, and more specifically, the compensatory gait strategies of patients with knee OA may offer yet another explanation for our findings. Specifically, these gait strategies include, inter alia, contralateral cane use, increased step width, and increased lateral trunk lean [10,11]. Inasmuch as these gait strategies involved increased COM acceleration and repetitive weight-shifts in the frontal plane [12], it follows that lateral trunk control is a crucial correlate of gait performance. That said, because we did not formally document the gait strategies in our sample, our inferences are at best speculative and future studies should examine whether seated postural control interacted with the different gait strategies to influence gait performance. Our findings have implications. First, perhaps our most compelling finding is that seated postural control was an independent correlate of gait performance, after adjusting for the effects of conventional knee impairments. Hitherto, knee OA studies examining pelvic control have narrowly focused on the lateral hip muscles [30,31], and very little is said about the close functional associations between the trunk and the pelvis. Our findings, if replicated, indicate that seated postural control should be considered in the clinical assessment of patients with knee OA. Furthermore, addressing possible trunk control impairments – for example, improving trunk musculature performance [32] – as part of the overall OA management could potentially enhance physical function. Second, from the research viewpoint, our results are methodologically important. Specifically, by demonstrating the convergent validity of the seated COP measures with gait performance, our results support the repurposing of the Wii
417
Board – a cheaply available and portable gaming device – as a practicable and novel assessment tool of seated postural control. Indeed, the current paucity of knee OA research on trunk control is probably, in part, a reflection of the logistical and technical difficulties of assessing seated postural control. For this reason, if our results were replicated, we envisage the potential application of the Wii Board in both clinical and research protocols. Our study has limitations in addition to those already stated above. First, our results may be affected by reverse causation due to their cross-sectional nature and intervention studies are clearly needed to establish the presence and direction of causation. Second, we studied patients with end-stage knee OA and it is uncertain if our findings are invariantly applicable to patients with mild to moderate knee OA. Third, we measured only gait performance, and it would be informative to evaluate other functional tasks – for example, stair climbing and sit-to-stand – given their potential abilities to tax trunk control in multiple planes. Relatedly, it would also be desirable to evaluate trunk control using an unstable sitting protocol given its ability to provide greater postural challenge to the trunk [7,16]. 5. Conclusions In patients with symptomatic knee OA, seated postural control – particularly in the ML direction – was associated with gait speed performance. Although our findings are strictly correlational, they suggest that seated postural control is an important correlate of physical function and that greater emphasis in the assessment of trunk control may be warranted. Finally, intervention studies are necessary to determine whether improvements in trunk control could enhance gait performance in patients with knee OA. Acknowledgements We thank our patients for donating their time. We also thank our colleagues, Bee-Yee Tan, Hanniel Lim, Felicia Seet, Jennifer Liaw, Hwei-Chi Chong, and William Yeo, from the Singapore General Hospital, for supporting this study. Finally, we thank the orthopaedic surgeons from the Singapore General Hospital for allowing us access to their patients. Conflict of interest None. References [1] Guccione AA, Felson DT, Anderson JJ, Anthony JM, Zhang YQ, Wilson PWF, et al. The effects of specific medical conditions on the functional limitations of elders in the Framingham Study. American Journal of Public Health 1994;84:351–8. [2] Sowers M, Jannausch ML, Gross M, Karvonen-Gutierrez CA, Palmieri RM, Crutchfield M, et al. Performance-based physical functioning in African-American and Caucasian women at midlife: considering body composition, quadriceps strength, and knee osteoarthritis. American Journal of Epidemiology 2006;163:950–8. [3] Schmitt LC, Rudolph KS. Influences on knee movement strategies during walking in persons with medial knee osteoarthritis. Arthritis and Rheumatism 2007;57:1018–26. [4] Maly MR, Costigan PA, Olney SJ. Role of knee kinematics and kinetics on performance and disability in people with medial compartment knee osteoarthritis. Clinical Biomechanics (Bristol Avon) 2006;21:1051–9. [5] Cholewicki J, Polzhofer GK, Radebold A. Postural control of trunk during unstable sitting. Journal of Biomechanics 2000;33:1733–7. [6] Bennett BC, Abel MF, Granata KP. Seated postural control in adolescents with idiopathic scoliosis. Spine (Philadelphia PA 1976) 2004;29:E449–54. [7] van Dieen JH, Koppes LL, Twisk JW. Postural sway parameters in seated balancing; their reliability and relationship with balancing performance. Gait and Posture 2010;31:42–6. [8] Winter DA. Biomechanics and motor control of human movement, 4th ed., New York: John Wiley & Sons Inc.; 2009. [9] Kang HG, Dingwell JB. Dynamic stability of superior vs. inferior segments during walking in young and older adults. Gait and Posture 2009;30:260–3.
418
Y.-H. Pua et al. / Gait & Posture 37 (2013) 413–418
[10] Hunt MA, Wrigley TV, Hinman RS, Bennell KL. Individuals with severe knee osteoarthritis (OA) exhibit altered proximal walking mechanics compared with individuals with less severe OA and those without knee pain. Arthritis Care and Research (Hoboken) 2010;62:1426–32. [11] Mundermann A, Asay JL, Mundermann L, Andriacchi TP. Implications of increased medio-lateral trunk sway for ambulatory mechanics. Journal of Biomechanics 2008;41:165–70. [12] Wang TM, Hsu WC, Chang CF, Hu CC, Lu TW. Effects of knee osteoarthritis on body’s center of mass motion in older adults during level walking. Biomedical Engineering Applications Basis and Communications 2010; 22:205–12. [13] Kaya BK, Krebs DE, Riley PO. Dynamic stability in elders: momentum control in locomotor ADL. Journals of Gerontology Series A Biological Sciences and Medical Sciences 1998;53:M126–34. [14] Visser JE, Carpenter MG, van der Kooij H, Bloem BR. The clinical utility of posturography. Clinical Neurophysiology 2008;119:2424–36. [15] Adkin AL, Frank JS, Carpenter MG, Peysar GW. Postural control is scaled to level of postural threat. Gait and Posture 2000;12:87–93. [16] van der Burg JC, van Wegen EE, Rietberg MB, Kwakkel G, van Dieen JH. Postural control of the trunk during unstable sitting in Parkinson’s disease. Parkinsonism & Related Disorders 2006;12:492–8. [17] Ritter PL, Gonzalez VM, Laurent DD, Lorig KR. Measurement of pain using the visual numeric scale. Journal of Rheumatology 2006;33:574–80. [18] Clark RA, Bryant AL, Pua YH, McCrory P, Bennell K, Hunt M. Validity and reliability of the Nintendo Wii Balance Board for assessment of standing balance. Gait and Posture 2010;31:307–10. [19] Ismail AR, Asfour SS. Discrete wavelet transform: a tool in smoothing kinematic data. Journal of Biomechanics 1999;32:317–21. [20] Thiebaut C, Roques S. Time-scale and time-frequency analyses of irregularly sampled astronomical time series. EURASIP Journal on Advances in Signal Processing 2005;15:2486–99. [21] Chiviacowsky S, Wulf G, Wally R. An external focus of attention enhances balance learning in older adults. Gait and Posture 2010;32:572–5.
[22] Brown M, Sinacore DR, Binder EF, Kohrt WM. Physical and performance measures for the identification of mild to moderate frailty. Journals of Gerontology Series A Biological Sciences and Medical Sciences 2000;55:M350–5. [23] Bigelow KE, Berme N. Development of a protocol for improving the clinical utility of posturography as a fall-risk screening tool. Journals of Gerontology Series A Biological Sciences and Medical Sciences 2011;66:228. [24] Williams HG, McClenaghan BA, Dickerson J. Spectral characteristics of postural control in elderly individuals. Archives of Physical Medicine and Rehabilitation 1997;78:737–44. [25] van Dieen JH, Koppes LL, Twisk JW. Low back pain history and postural sway in unstable sitting. Spine (Philadelphia PA 1976) 2010;35:812–7. [26] Bauby CE, Kuo AD. Active control of lateral balance in human walking. Journal of Biomechanics 2000;33:1433–40. [27] McAndrew PM, Dingwell JB, Wilken JM. Walking variability during continuous pseudo-random oscillations of the support surface and visual field. Journal of Biomechanics 2010;43:1470–5. [28] McAndrew PM, Wilken JM, Dingwell JB. Dynamic stability of human walking in visually and mechanically destabilizing environments. Journal of Biomechanics 2011;44:644–9. [29] Hurt CP, Rosenblatt N, Crenshaw JR, Grabiner MD. Variation in trunk kinematics influences variation in step width during treadmill walking by older and younger adults. Gait and Posture 2010;31:461–4. [30] Sled EA, Khoja L, Deluzio KJ, Olney SJ, Culham EG. Effect of a home program of hip abductor exercises on knee joint loading, strength, function, and pain in people with knee osteoarthritis: a clinical trial. Physical Therapy 2010;90:895–904. [31] Bennell KL, Hunt MA, Wrigley TV, Hunter DJ, McManus FJ, Hodges PW, et al. Hip strengthening reduces symptoms but not knee load in people with medial knee osteoarthritis and varus malalignment: a randomised controlled trial. Osteoarthritis and Cartilage 2010;18:621–8. [32] Suri P, Kiely DK, Leveille SG, Frontera WR, Bean JF. Increased trunk extension endurance Is associated with meaningful improvement in balance among older adults with mobility problems. Archives of Physical Medicine and Rehabilitation 2011;92:1038–43.