JCLB-03927; No of Pages 4 Clinical Biomechanics xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults Boon-Chong Kwok a, Ross A. Clark b, Yong-Hao Pua c,⁎ a b c
Clinical Services (Collaborative Care), National Healthcare Group Polyclinics, 3 Fusionpolis Link, Nexus@one-north, Singapore School of Exercise Science, Australian Catholic University, Melbourne, Australia Department of Physiotherapy, Singapore General Hospital, Outram Road, Singapore
a r t i c l e
i n f o
Article history: Received 31 July 2014 Accepted 4 March 2015 Keywords: Balance Older adult Primary care Physiotherapy
a b s t r a c t Background: The Wii Balance Board has received increasing attention as a balance measurement tool; however its ability to prospectively predict falls is unknown. This exploratory study investigated the use of the Wii Balance Board and other clinical-based measures for prospectively predicting falls among community-dwelling older adults. Methods: Seventy-three community-dwelling men and women, aged 60–85 years were followed-up over a year for falls. Standing balance was indexed by sway velocities measured using the Wii Balance Board interfaced with a laptop. Clinical-based measures included Short Physical Performance Battery, gait speed and Timed-Up-and-Go test. Multivariable regression analyses were used to assess the ability of the Wii Balance Board measure to complement the TUG test in fall screening. Findings: Individually, the study found Wii Balance Board anteroposterior (odds ratio 1.98, 95% CI 1.16 to 3.40, P = 0.01) and mediolateral (odds ratio 2.80, 95% CI 1.10 to 7.13, p = 0.03) sway velocity measures predictive of prospective falls. However, when each velocity measure was adjusted with body mass index and Timed-Upand-Go, only anteroposterior sway velocity was predictive of prospective falls (odds ratio 2.21, 95% CI 1.18 to 4.14). A faster anteroposterior velocity was associated with increased odds of falling. Area-under-the-curves for Wii Balance Board sway velocities were 0.67 and 0.71 for anteroposterior and mediolateral respectively. Interpretation: The Wii Balance Board-derived anteroposterior sway velocity measure could complement existing clinical-based measures in predicting future falls among community-dwelling older adults. Trial registration: Australian New Zealand Clinical Trials Registry number: ACTRN12610001099011. © 2015 Elsevier Ltd. All rights reserved.
1. Background Accidental falls and fall-related injuries among older people are important health issues (Herala et al., 2000). A fall can occur due to balance impairment, which can be averted with early detection (Bigelow and Berme, 2011; Whitney et al., 2005). There are currently two methods of balance assessments to predict falls, clinical-based and laboratorybased balance tests (Botolfsen et al., 2008; Close and Lord, 2011; Mancini and Horak, 2010; Pajala et al., 2008). In contrast to laboratory-based balance tests, most clinical-based measures do not test balance as a single construct (Botolfsen et al., 2008; Close and Lord, 2011; Guralnik et al., 1994). Indeed, among the Abbreviations: WBB, Wii Balance Board; CoP, center of pressure; AP, anterior–posterior; ML, medial-lateral; USD,UnitedStates dollars; MDC, minimaldetectablechange; SPPB, short physical performance battery; TUG, Timed-Up-and-Go; OR, odds ratio; CI, confidence interval. ⁎ Corresponding author at: Department of Physiotherapy, Singapore General Hospital, Outram Road, 169608 Singapore. E-mail addresses: [email protected] (B.-C. Kwok), [email protected] (R.A. Clark), [email protected] (Y.-H. Pua).
clinical-based measures, the Timed-Up-and-Go (TUG) test involves myriad components of reaction time, lower limb strength, gait speed and agility to evaluate function and predict falls (Botolfsen et al., 2008; Close and Lord, 2011; Mancini and Horak, 2010). Another form of clinical-based measure, the short physical performance battery (SPPB), provides an evaluation of functioning status with components of balance, gait speed and lower limb endurance tests (Close and Lord, 2011; Freire et al., 2012; Guralnik et al., 1994). Despite the multiconstruct nature of the clinical-based tests, the TUG and other clinicalbased measures have only low to moderate predictive validity for falls among older adults (Lee et al., 2013; Power et al., 2014). To complement existing clinical-based measures in predicting falls, balance is best evaluated with the laboratory-based force-platederived center of pressure (CoP) measure (Bigelow and Berme, 2011; Pajala et al., 2008). However, the force-plate is neither small nor practical enough to be applied in large community cohorts (Pajala et al., 2008; Schneider et al., 2011). The portable gaming Nintendo Wii Balance Board (WBB) could potentially serve as a bridge between clinicalbased and force-plate balance tests. Previously, the WBB was used to evaluate standing CoP (Clark et al., 2010), while a recent study found
http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006 0268-0033/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
2
B.-C. Kwok et al. / Clinical Biomechanics xxx (2015) xxx–xxx
that the WBB CoP measure was associated with fear-induced activity limitation (Pua et al., 2013), but the validity of the WBB for prospectively predicting falls has not been established. This pilot study was primarily undertaken to investigate the WBB predictive validity for prospective falls. Second, this study investigated the WBB CoP measure to complement existing clinical-based measures to predict falls for the community-dwelling older adults. 2. Methods 2.1. Participants The study sample was derived from the randomized controlled trial “Evaluation of the Frails' Fall Efficacy by Comparing Treatments” (EFFECT) study and detailed recruitment inclusion and exclusion criteria have been described elsewhere (Kwok et al., 2011). The study included 73 community-dwelling older adults above 60 years old and participants were primarily excluded for significant cognitive disorder (Abbreviated Mental Test score below 7 for 60–74 years old, below 6 for 75 years old and above), and unstable medical or surgical conditions (Kwok et al., 2011). This study was undertaken from week 13 in the EFFECT study. The participants were assessed with clinical-based measures (TUG and SPPB) and WBB CoP velocity measures. Gait speed was calculated from the 4-meter walk of SPPB. The study was approved by the Centralised Institutional Review Board (Reference: 2010/177/D and 2010/639/D) and written informed consent was obtained from the participants prior to the study enrolment. No adverse events were reported in this study. 2.2. Clinical-based measures TUG test: The test required participant to rise from a chair (seat height 45 cm) without using the arms to assist, walk 3 m to a cone, turn around the cone and return to the seat (Shumway-Cook et al., 2000). Participants were instructed to walk safely with prior demonstration (Podsiadlo and Richardson, 1991). This test has good test–retest reliability, intra-class correlation coefficient (ICC) 0.84 (Botolfsen et al., 2008). SPPB: This measure comprises three physical tests — tandem stance, gait speed and 5 times chair stands (Guralnik et al., 1994). These validated tests provide useful information about physical frailty (functional decline) of an older adult and have good test–retest reliability, ICC 0.83 to 0.89 (Freire et al., 2012). 2.3. WBB CoP measures Standing balance was assessed using the WBB (Nintendo, Kyoto, Japan) in a protocol previously validated against a laboratory forceplate (Clark et al., 2010). To perform the test, participants stood barefooted on the WBB in their usual comfortable stance, and they were instructed to keep their hands by their side, look straight ahead, and stand quietly. A previous study showed that this safe assessment method had similar fall prediction accuracy when compared with more challenging tasks such as eyes closed tandem stance (Pajala et al., 2008). Two 30-second trials were performed, and the mean of two trials was used. The WBB was interfaced with a laptop computer using customwritten software (Labview 8.5 National Instruments, Austin, TX, U.S.A.), and was calibrated by placing a variety of known loads at different positions on the WBB. AP and ML CoP coordinates were recorded at 40 Hz and low-pass filtered at 6.25 Hz using an undecimated Symlet-8 wavelet with the detail levels removed. Given the multitude of CoP measures, we have a priori focused on AP and ML sway velocity to minimize Type I error from multiplicity (Ottenbacher, 1998). CoP velocity
represents the sway distance covered by the CoP in the respective planes (path length) divided by the sampling duration (30 s). 2.4. Prospective falls The primary outcome was prospective falls over a year using Kellogg's definition of a fall: “a fall is an event which results in a person coming to rest inadvertently on the ground or other lower level and other than as a consequence of the following: sustaining a violent blow, loss of consciousness, sudden onset of paralysis as in a stroke, or an epileptic seizure.”Zecevic et al., 2006 In order to minimize recall bias, we employed three methods in collecting fall data. Firstly, participants were instructed to record events of falling during the one-year follow-up by indicating their occurrence on a fall calendar. Second, they were contacted monthly via telephone to check for any event of fall (Fleming et al., 2008). Last, the study Principal Investigator initiated a 24-hour point of contact for participants to report falls. 2.5. Statistical analysis We used descriptive statistics to characterize the study sample: means with SDs for continuous variables and frequencies with percentages for categorical variables. We analyzed the characteristics of age, body mass index (BMI), gender, ethnicity, abbreviated mental test score, medical co-morbidities, non-psychotropic medication usage per day, visual contrast and 1-year fall history, between non-fallers and fallers with Student's t-tests or Pearson's chi-squared tests. Fallers and non-fallers were similarly compared on SPPB, TUG, and gait speed measures, and WBB velocity measures using Student's t-test. In addition, Spearman's correlation was used to establish the association between SPPB, TUG, gait speed, and WBB velocity measures. The association between fall outcome and each measure was first analyzed using univariable logistic regression. Next, the multivariable logistic regression model, adjusted for BMI (a standard covariate associated with falls) (Sheehan et al., 2013), was used to investigate if WBB AP or ML velocity measure could complement the TUG measure — a common clinical-based measure (Close and Lord, 2011). Because we used data from a treatment study, treatment variable and its interaction with the screening measures were also included in the multivariable models. As the joint test for treatment and its interaction terms gave non-significant results, we removed these terms from the final model. Given that the various measures were quantified on different scales, the ORs were scaled to the difference between the 75th and the 25th percentiles (the interquartile range) of each measure (Harrell, 2001). Besides allowing valid between-measure comparisons, interquartile range ORs represented a more clinically meaningful distinction than the conventional one-unit change in predictor values (Harrell, 2001). The area under the receiver-operating characteristic curve (AUC) was used to assess the predictive validity of falls for each measure. An AUC value of 1 represents perfect discrimination and 0.5 represents chance discrimination. All statistical analyses were performed with SPSS version 19.0 (Chicago: SPSS Inc.). Statistical significance was set at P(2-tailed) b 0.05. 3. Results In this study, 18 (~25%) participants reported a fall in the 12-month monitoring period. The demographics of the non-fallers and fallers are presented in Table 1, which showed no statistical difference between the two groups except for WBB velocity measures (P b 0.05). All the clinical-based measures were weakly correlated to the WBB velocity measures (Table 2). The study found that at the univariable level, older adults had increased odds of falling with higher WBB velocity measures (ORs ≥ 1.98, P's b 0.05), but the clinical-based measures did not reach statistical significance (Table 3). Both the AUC values for the WBB AP
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
B.-C. Kwok et al. / Clinical Biomechanics xxx (2015) xxx–xxx Table 1 Characteristics of non-fallers (n = 55) and fallers (n = 18). Characteristics
Non-fallers
Age (years) Body mass index (kg m−2) Gender — Female, n (%) Ethnic — Chinese, n (%) Abbreviated Mental Test score (/10) Medical history — co-morbidities, n (%) 0 1 2 ≥3
69.7 (7.8) 22.8 (3.7) 45 (81.8) 54 (98.2) 8.9 (1.1) 6 (10.9) 10 (18.2) 15 (27.2) 24 (43.6)
3
Table 3 Predictive validity of each screening measure for future falls. Fallers
OR (95% CI)a
P value
AUC (95% CI)
0.62 (0.36 to 1.10)
0.10
0.61 (0.46 to 0.77)
1.31 (0.90 to 1.96) 0.51 (0.22 to 1.25)
0.16 0.14
0.64 (0.50 to 0.79) 0.61 (0.46 to 0.77)
0.01
0.67 (0.52 to 0.82)
0.03
0.71 (0.56 to 0.85)
P value
Screening measures
70.7 (5.2) 21.6 (3.7) 17 (94.4) 17 (94.4) 8.7 (1.3)
0.64 0.26 0.20 0.40 0.33
Clinical balance test Short Physical Performance Battery score (/12) Timed-Up-and-Go (s) Gait speed (m s−1)
0 (0) 4 (22.2) 6 (33.3) 8 (44.4)
0.83
Wii Balance Board test (unadjusted) 1.98 (1.16 to 3.40) Anteroposterior velocity (cm s−1) −1 Mediolateral velocity (cm s ) 2.80 (1.10 to 7.13) a
Medications per day, n (%) 0 1 2 ≥3 Vision — Melbourne-edge test (dB) Had a fall in the past 1 year, n (%) Screening measures Short Physical Performance Battery score (/12) Timed-Up-and-Go (s) Gait speed (ms−1) WBB anteroposterior velocity (cm s−1) WBB mediolateral velocity (cm s−1)
14 (25.5) 7 (12.7) 7 (12.7) 27 (49.1) 19.7 (1.6) 28 (51.0)
3 (16.7) 4 (22.2) 2 (11.1) 9 (50.0) 18.9 (2.5) 13 (72.2)
0.61
0.19 0.12
10.6 (1.6)
9.8 (2.0)
0.14
9.10 (4.07) 1.27 (0.44) 1.02 (0.26) 0.73 (0.13)
10.91 (5.47) 1.10 (0.42) 1.27 (0.45) 0.82 (0.14)
0.14 0.14 0.049 0.02
Continuous variables were analyzed with t-test and presented as mean (SD). Categorical variables were analyzed with chi-squared test. WBB: Wii Balance Board.
and ML velocity measures were statistically significant: 0.67 and 0.71 respectively (P b 0.05). At the multivariable level (Table 4), only WBB AP velocity retained its predictive ability of falls (OR 2.21, 95% CI 1.18 to 4.14, P = 0.01), whereas WBB ML velocity did not (OR 2.43, 95% CI 0.86 to 6.90, P = 0.09). 4. Discussion This prospective pilot study identified and established the predictive validity of WBB-derived CoP velocity measures for falls. We found that the WBB balance measure – in particular, the velocity AP measure – could potentially complement the TUG to predict future falls in community-dwelling older adults. 4.1. Relationship between clinical-based and WBB measures Each of the clinical-based measures was moderately correlated to each other but not with the WBB measures (Table 2). The weak relationship between the clinical-based measure and the WBB balance measure showed that the two methods were non-redundant with one another. Our study finding was similar to that of recent literature investigating the relationship between clinical-based and laboratory balance measures (Nguyen et al., 2012). The main difference between these two methods of measures was that the WBB could quantify gross to subtle sway displacements, while clinical-based measures could not identify Table 2 Non-parametric correlation of screening measures.
ORs estimate the odds of future falls at the 75th vs the 25th percentile for all continuous predictors. For example, increasing the mediolateral velocity variable from its lower quartile (0.65 cm s−1) to its higher quartile (0.84 cm s−1) was associated with a 2.8-fold (95%CI, 1.10- to 7.13-fold) increase in the odds of future falls.
by visual inspection (Mancini and Horak, 2010). Thus, in community fall screening, the WBB balance measure could provide valuable information to complement existing clinical-based measures to identify individuals with fall risk. 4.2. Clinical-based measures The clinical-based measures did not predict falls among communitydwelling older adults. Despite a combination of balance, gait speed and lower limb strength measures, the SPPB was not predictive of falls. The TUG measure was also not predictive of falls in our study, similar to previous literature (Boulgarides et al., 2003; Haines et al., 2008), of which one study attributed the null findings to a possible ceiling effect in higher-functioning older adults (Boulgarides et al., 2003). Higherfunctioning older adults are at risk of fall but would not be identified based on cut-off values of 14 s in one study and 15 s in another (Shumway-Cook et al., 2000; Whitney et al., 2005). Finally, as suggested by the Reviewer, another potential explanation for the null results is that because clinical-based measures are multi-construct in nature, their “non-predictive” components may attenuate the overall predictive value of the test. Future studies should examine this possibility. 4.3. WBB balance measure WBB CoP sway velocity quantifies the speed of the movement required to maintain postural stability, and reflects the postural response of the neuromuscular system to somatosensory, visual and vestibular information. Older adults with poorer balance may produce greater muscle co-contraction or stiffening to maintain balance (Carpenter et al., 2001; Ho and Bendrups, 2002), which would result in rapid movements, and hence conceivably higher CoP velocities. Irrespective of AP or ML directions in this study, the WBB velocity measures had better odds in predicting future falls (Table 3) and these findings concur with existing evidence using laboratory force-plate to identify fallers (Bigelow and Berme, 2011; Schneider et al., 2011). However, when adjusted for BMI and in complementing the TUG measure (Table 4), only the AP velocity measure remained predictive of falls. In other words, the predictive effects of AP velocity for future falls were not confounded by those of TUG and BMI; hence, AP velocity could complement the TUG in fall screening. Earlier literature had shown that the AP velocity was higher among older adults with impaired balance (Baloh et al., 1995; Pajala et al., 2008). 4.4. Study implications and limitation
TUG: Timed-Up-and-Go, SPPB: Short Physical Performance Battery, WBB: Wii Balance Board, AP: anteroposterior, ML: mediolateral. ⁎ P b 0.001. ⁎⁎ P b 0.05.
The study provided the first evidence for the WBB-derived CoP velocity measures to predict future falls in a sample of communitydwelling older adults. All participants completed two 30-second standing trials on the WBB without adverse events, supporting the time-
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
4
B.-C. Kwok et al. / Clinical Biomechanics xxx (2015) xxx–xxx
and effort, and the co-authors of previously published literature that led to this study.
Table 4 Logistic regression adjusted for BMI and TUG to predict future falls. Variables
25th percentile
75th percentile
OR (95% CI)a
P value
BMI (kg m−2) TUG (s) AP velocity (cm s−1)
20.2 6.75 0.86
24.6 10.3 1.16
0.51 (0.22 to 1.16) 1.24 (0.81 to 1.91) 2.21 (1.18 to 4.14)
0.11 0.32 0.01
BMI (kg m−2) TUG (s) ML velocity (cm s−1)
20.2 6.75 0.65
24.6 10.3 0.84
0.89 (0.40 to 1.95) 1.27 (0.83 to 1.94) 2.43 (0.86 to 6.90)
0.76 0.28 0.09
References
BMI: body mass index, TUG: Timed-Up-and-Go, AP: anteroposterior, ML: mediolateral. a ORs scaled by the respective screening measure interquartile range.
efficiency and feasibility of WBB testing in field applications. To the clinicians, the WBB screening tool potentially provides a quick and valid measure of balance and fall risk; to the health policy-makers, the WBB may be a technologically advanced yet an affordable screening investment in primary care (WBB and a laptop/tablet combined cost b$500USD) as compared to laboratory-based force plate (more than $10,000USD). Our study has limitations. One study limitation was the lack of home exercise adherence monitoring in the prospective 1-year follow-up, which might have influenced the fall incidence. Another limitation is that during WBB testing, we evaluated bipedal standing so that even older adults with substantial balance limitations could perform the test safely. Nevertheless, it may be informative to evaluate other balance tasks, for example, tandem stances on the WBB, given their abilities to provide greater challenge to balance control. A final limitation of the explorative study was the low sample size. 5. Conclusion The present study supported the predictive validity of the WBB measures for future falls. Specifically, the WBB-derived CoP measures – in particular, the velocity AP measure – could potentially complement the TUG to predict future falls in community-dwelling older adults. Authors' contributions All authors participated in the study design. RAC created the data collection and analysis system. BCK carried out the data collection. YHP and BCK performed the statistical analysis. BCK drafted the manuscript. All authors provided input into the manuscript draft and have read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Acknowledgments The SingHealth Foundation research grant (Grant number: SHF/ FG397S/2009), the Singapore General Hospital research grant (Grant number: SRG04/2010) and the Singapore Physiotherapy Association research grant (Grant number: RF09-005) funded the study. The funders were not involved in the study. We would like to thank Ms Prithvi Balaji (Physiotherapist) for her role as an outcome assessor in the EFFECT study, Mr Kwok Boon Yong (Lecturer) from the School of Communication, Arts and Social Sciences, Singapore Polytechnic for providing language and editorial input, the participants who volunteered their time
Baloh, R.W., Spain, S., Socotch, T.M., Jacobson, K.M., Bell, T., 1995. Posturography and balance problems in older people. J. Am. Geriatr. Soc. 43, 638–644. Bigelow, K.E., Berme, N., 2011. Development of a protocol for improving the clinical utility of posturography as a fall-risk screening tool. J. Gerontol. A Biol. Sci. Med. Sci. 66, 228–233. Botolfsen, P., Helbostad, J.L., Moe-Nilssen, R., Wall, J.C., 2008. Reliability and concurrent validity of the Expanded Timed Up-and-Go test in older people with impaired mobility. Physiother. Res. Int. 13, 94–106. Boulgarides, L.K., McGinty, S.M., Willett, J.A., Barnes, C.W., 2003. Use of clinical and impairment-based tests to predict falls by community-dwelling older adults. Phys. Ther. 83, 328–339. Carpenter, M.G., Frank, J.S., Silcher, C.P., Peysar, G.W., 2001. The influence of postural threat on the control of upright stance. Exp. Brain Res. 138, 210–218. Clark, R.A., Bryant, A.L., Pua, Y., McCrory, P., Bennell, K., Hunt, M., 2010. Validity and reliability of the Nintendo Wii Balance Board for assessment of standing balance. Gait Posture 31, 307–310. Close, J.C., Lord, S.R., 2011. Fall assessment in older people. BMJ 343, d5153. Fleming, J., Matthews, F.E., Brayne, C., 2008. Falls in advanced old age: recalled falls and prospective follow-up of over-90-year-olds in the Cambridge City over-75's Cohort study. BMC Geriatr. 8, 6. Freire, A.N., Guerra, R.O., Alvarado, B., Guralnik, J.M., Zunzunegui, M.V., 2012. Validity and reliability of the short physical performance battery in two diverse older adult populations in Quebec and Brazil. J. Aging Health 24, 863–878. Guralnik, J.M., Simonsick, E.M., Ferrucci, L., et al., 1994. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J. Gerontol. 49, M85–M94. Haines, T., Kuys, S.S., Morrison, G., Clarke, J., Bew, P., 2008. Balance impairment not predictive of falls in geriatric rehabilitation wards. J. Gerontol. A Biol. Sci. Med. Sci. 63, 523–528. Harrell Jr., F.E., 2001. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. Springer, New York. Herala, M., Luukinen, H., Honkanen, R., Koski, K., Laippala, P., Kivela, S.L., 2000. Soft tissue injury resulting from falling predicts a future major falling injury in the home dwelling elderly. J. Epidemiol. Community Health 54, 557–558. Ho, C.Y., Bendrups, A.P., 2002. Ankle reflex stiffness during unperceived perturbation of standing in elderly subjects. J. Gerontol. A Biol. Sci. Med. Sci. 57, B344–B350. Kwok, B.C., Mamun, K., Chandran, M., Wong, C.H., 2011. Evaluation of the Frails' Fall Efficacy by Comparing Treatments (EFFECT) on reducing fall and fear of fall in moderately frail older adults: study protocol for a randomised control trial. Trials 12, 155. Lee, J., Geller, A.I., Strasser, D.C., 2013. Analytical review: focus on fall screening assessments. PM R 5, 609–621. Mancini, M., Horak, F.B., 2010. The relevance of clinical balance assessment tools to differentiate balance deficits. Eur. J. Phys. Rehabil. Med. 46, 239–248. Nguyen, U.S., Kiel, D.P., Li, W., et al., 2012. Correlations of clinical and laboratory measures of balance in older men and women. Arthritis Care Res. 64, 1895–1902. Ottenbacher, K.J., 1998. Quantitative evaluation of multiplicity in epidemiology and public health research. Am. J. Epidemiol. 147, 615–619. Pajala, S., Era, P., Koskenvuo, M., Kaprio, J., Tormakangas, T., Rantanen, T., 2008. Force platform balance measures as predictors of indoor and outdoor falls in communitydwelling women aged 63–76 years. J. Gerontol. A Biol. Sci. Med. Sci. 63, 171–178. Podsiadlo, D., Richardson, S., 1991. The timed “Up & Go”: a test of basic functional mobility for frail elderly persons. J. Am. Geriatr. Soc. 39, 142–148. Power, V., Van de Ven, P., Nelson, J., Clifford, A.M., 2014. Predicting falls in communitydwelling older adults: a systematic review of task performance-based assessment tools. Physiother. Pract. Res. 35, 3–15. Pua, Y.H., Ong, P.H., Lee, A.Y., Tan, J., Bryant, A.L., Clark, R.A., 2013. Preliminary prediction model for fear-induced activity limitation after total knee arthroplasty in people 60 years and older: prospective cohort study. Arch. Phys. Med. Rehabil. 94, 503–509. Schneider, P., Schwab, M., Hanscheid, H., 2011. Identification of factors associated with risk of fall using a force platform and power spectrum analysis technique. J. Biomech. 44, 2008–2012. Sheehan, K.J., O'Connell, M.D., Cunningham, C., Crosby, L., Kenny, R.A., 2013. The relationship between increased body mass index and frailty on falls in community dwelling older adults. BMC Geriatr. 13, 132. Shumway-Cook, A., Brauer, S., Woollacott, M., 2000. Predicting the probability for falls in community-dwelling older adults using the Timed Up & Go Test. Phys. Ther. 80, 896–903. Whitney, J.C., Lord, S.R., Close, J.C., 2005. Streamlining assessment and intervention in a falls clinic using the Timed Up and Go Test and physiological profile assessments. Age Ageing 34, 567–571. Zecevic, A.A., Salmoni, A.W., Speechley, M., Vandervoort, A.A., 2006. Defining a fall and reasons for falling: comparisons among the views of seniors, health care providers, and the research literature. The Gerontologist 46, 367–376.
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults Boon-Chong Kwok a, Ross A. Clark b, Yong-Hao Pua c,⁎ a b c
Clinical Services (Collaborative Care), National Healthcare Group Polyclinics, 3 Fusionpolis Link, Nexus@one-north, Singapore School of Exercise Science, Australian Catholic University, Melbourne, Australia Department of Physiotherapy, Singapore General Hospital, Outram Road, Singapore
a r t i c l e
i n f o
Article history: Received 31 July 2014 Accepted 4 March 2015 Keywords: Balance Older adult Primary care Physiotherapy
a b s t r a c t Background: The Wii Balance Board has received increasing attention as a balance measurement tool; however its ability to prospectively predict falls is unknown. This exploratory study investigated the use of the Wii Balance Board and other clinical-based measures for prospectively predicting falls among community-dwelling older adults. Methods: Seventy-three community-dwelling men and women, aged 60–85 years were followed-up over a year for falls. Standing balance was indexed by sway velocities measured using the Wii Balance Board interfaced with a laptop. Clinical-based measures included Short Physical Performance Battery, gait speed and Timed-Up-and-Go test. Multivariable regression analyses were used to assess the ability of the Wii Balance Board measure to complement the TUG test in fall screening. Findings: Individually, the study found Wii Balance Board anteroposterior (odds ratio 1.98, 95% CI 1.16 to 3.40, P = 0.01) and mediolateral (odds ratio 2.80, 95% CI 1.10 to 7.13, p = 0.03) sway velocity measures predictive of prospective falls. However, when each velocity measure was adjusted with body mass index and Timed-Upand-Go, only anteroposterior sway velocity was predictive of prospective falls (odds ratio 2.21, 95% CI 1.18 to 4.14). A faster anteroposterior velocity was associated with increased odds of falling. Area-under-the-curves for Wii Balance Board sway velocities were 0.67 and 0.71 for anteroposterior and mediolateral respectively. Interpretation: The Wii Balance Board-derived anteroposterior sway velocity measure could complement existing clinical-based measures in predicting future falls among community-dwelling older adults. Trial registration: Australian New Zealand Clinical Trials Registry number: ACTRN12610001099011. © 2015 Elsevier Ltd. All rights reserved.
1. Background Accidental falls and fall-related injuries among older people are important health issues (Herala et al., 2000). A fall can occur due to balance impairment, which can be averted with early detection (Bigelow and Berme, 2011; Whitney et al., 2005). There are currently two methods of balance assessments to predict falls, clinical-based and laboratorybased balance tests (Botolfsen et al., 2008; Close and Lord, 2011; Mancini and Horak, 2010; Pajala et al., 2008). In contrast to laboratory-based balance tests, most clinical-based measures do not test balance as a single construct (Botolfsen et al., 2008; Close and Lord, 2011; Guralnik et al., 1994). Indeed, among the Abbreviations: WBB, Wii Balance Board; CoP, center of pressure; AP, anterior–posterior; ML, medial-lateral; USD,UnitedStates dollars; MDC, minimaldetectablechange; SPPB, short physical performance battery; TUG, Timed-Up-and-Go; OR, odds ratio; CI, confidence interval. ⁎ Corresponding author at: Department of Physiotherapy, Singapore General Hospital, Outram Road, 169608 Singapore. E-mail addresses: [email protected] (B.-C. Kwok), [email protected] (R.A. Clark), [email protected] (Y.-H. Pua).
clinical-based measures, the Timed-Up-and-Go (TUG) test involves myriad components of reaction time, lower limb strength, gait speed and agility to evaluate function and predict falls (Botolfsen et al., 2008; Close and Lord, 2011; Mancini and Horak, 2010). Another form of clinical-based measure, the short physical performance battery (SPPB), provides an evaluation of functioning status with components of balance, gait speed and lower limb endurance tests (Close and Lord, 2011; Freire et al., 2012; Guralnik et al., 1994). Despite the multiconstruct nature of the clinical-based tests, the TUG and other clinicalbased measures have only low to moderate predictive validity for falls among older adults (Lee et al., 2013; Power et al., 2014). To complement existing clinical-based measures in predicting falls, balance is best evaluated with the laboratory-based force-platederived center of pressure (CoP) measure (Bigelow and Berme, 2011; Pajala et al., 2008). However, the force-plate is neither small nor practical enough to be applied in large community cohorts (Pajala et al., 2008; Schneider et al., 2011). The portable gaming Nintendo Wii Balance Board (WBB) could potentially serve as a bridge between clinicalbased and force-plate balance tests. Previously, the WBB was used to evaluate standing CoP (Clark et al., 2010), while a recent study found
http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006 0268-0033/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
2
B.-C. Kwok et al. / Clinical Biomechanics xxx (2015) xxx–xxx
that the WBB CoP measure was associated with fear-induced activity limitation (Pua et al., 2013), but the validity of the WBB for prospectively predicting falls has not been established. This pilot study was primarily undertaken to investigate the WBB predictive validity for prospective falls. Second, this study investigated the WBB CoP measure to complement existing clinical-based measures to predict falls for the community-dwelling older adults. 2. Methods 2.1. Participants The study sample was derived from the randomized controlled trial “Evaluation of the Frails' Fall Efficacy by Comparing Treatments” (EFFECT) study and detailed recruitment inclusion and exclusion criteria have been described elsewhere (Kwok et al., 2011). The study included 73 community-dwelling older adults above 60 years old and participants were primarily excluded for significant cognitive disorder (Abbreviated Mental Test score below 7 for 60–74 years old, below 6 for 75 years old and above), and unstable medical or surgical conditions (Kwok et al., 2011). This study was undertaken from week 13 in the EFFECT study. The participants were assessed with clinical-based measures (TUG and SPPB) and WBB CoP velocity measures. Gait speed was calculated from the 4-meter walk of SPPB. The study was approved by the Centralised Institutional Review Board (Reference: 2010/177/D and 2010/639/D) and written informed consent was obtained from the participants prior to the study enrolment. No adverse events were reported in this study. 2.2. Clinical-based measures TUG test: The test required participant to rise from a chair (seat height 45 cm) without using the arms to assist, walk 3 m to a cone, turn around the cone and return to the seat (Shumway-Cook et al., 2000). Participants were instructed to walk safely with prior demonstration (Podsiadlo and Richardson, 1991). This test has good test–retest reliability, intra-class correlation coefficient (ICC) 0.84 (Botolfsen et al., 2008). SPPB: This measure comprises three physical tests — tandem stance, gait speed and 5 times chair stands (Guralnik et al., 1994). These validated tests provide useful information about physical frailty (functional decline) of an older adult and have good test–retest reliability, ICC 0.83 to 0.89 (Freire et al., 2012). 2.3. WBB CoP measures Standing balance was assessed using the WBB (Nintendo, Kyoto, Japan) in a protocol previously validated against a laboratory forceplate (Clark et al., 2010). To perform the test, participants stood barefooted on the WBB in their usual comfortable stance, and they were instructed to keep their hands by their side, look straight ahead, and stand quietly. A previous study showed that this safe assessment method had similar fall prediction accuracy when compared with more challenging tasks such as eyes closed tandem stance (Pajala et al., 2008). Two 30-second trials were performed, and the mean of two trials was used. The WBB was interfaced with a laptop computer using customwritten software (Labview 8.5 National Instruments, Austin, TX, U.S.A.), and was calibrated by placing a variety of known loads at different positions on the WBB. AP and ML CoP coordinates were recorded at 40 Hz and low-pass filtered at 6.25 Hz using an undecimated Symlet-8 wavelet with the detail levels removed. Given the multitude of CoP measures, we have a priori focused on AP and ML sway velocity to minimize Type I error from multiplicity (Ottenbacher, 1998). CoP velocity
represents the sway distance covered by the CoP in the respective planes (path length) divided by the sampling duration (30 s). 2.4. Prospective falls The primary outcome was prospective falls over a year using Kellogg's definition of a fall: “a fall is an event which results in a person coming to rest inadvertently on the ground or other lower level and other than as a consequence of the following: sustaining a violent blow, loss of consciousness, sudden onset of paralysis as in a stroke, or an epileptic seizure.”Zecevic et al., 2006 In order to minimize recall bias, we employed three methods in collecting fall data. Firstly, participants were instructed to record events of falling during the one-year follow-up by indicating their occurrence on a fall calendar. Second, they were contacted monthly via telephone to check for any event of fall (Fleming et al., 2008). Last, the study Principal Investigator initiated a 24-hour point of contact for participants to report falls. 2.5. Statistical analysis We used descriptive statistics to characterize the study sample: means with SDs for continuous variables and frequencies with percentages for categorical variables. We analyzed the characteristics of age, body mass index (BMI), gender, ethnicity, abbreviated mental test score, medical co-morbidities, non-psychotropic medication usage per day, visual contrast and 1-year fall history, between non-fallers and fallers with Student's t-tests or Pearson's chi-squared tests. Fallers and non-fallers were similarly compared on SPPB, TUG, and gait speed measures, and WBB velocity measures using Student's t-test. In addition, Spearman's correlation was used to establish the association between SPPB, TUG, gait speed, and WBB velocity measures. The association between fall outcome and each measure was first analyzed using univariable logistic regression. Next, the multivariable logistic regression model, adjusted for BMI (a standard covariate associated with falls) (Sheehan et al., 2013), was used to investigate if WBB AP or ML velocity measure could complement the TUG measure — a common clinical-based measure (Close and Lord, 2011). Because we used data from a treatment study, treatment variable and its interaction with the screening measures were also included in the multivariable models. As the joint test for treatment and its interaction terms gave non-significant results, we removed these terms from the final model. Given that the various measures were quantified on different scales, the ORs were scaled to the difference between the 75th and the 25th percentiles (the interquartile range) of each measure (Harrell, 2001). Besides allowing valid between-measure comparisons, interquartile range ORs represented a more clinically meaningful distinction than the conventional one-unit change in predictor values (Harrell, 2001). The area under the receiver-operating characteristic curve (AUC) was used to assess the predictive validity of falls for each measure. An AUC value of 1 represents perfect discrimination and 0.5 represents chance discrimination. All statistical analyses were performed with SPSS version 19.0 (Chicago: SPSS Inc.). Statistical significance was set at P(2-tailed) b 0.05. 3. Results In this study, 18 (~25%) participants reported a fall in the 12-month monitoring period. The demographics of the non-fallers and fallers are presented in Table 1, which showed no statistical difference between the two groups except for WBB velocity measures (P b 0.05). All the clinical-based measures were weakly correlated to the WBB velocity measures (Table 2). The study found that at the univariable level, older adults had increased odds of falling with higher WBB velocity measures (ORs ≥ 1.98, P's b 0.05), but the clinical-based measures did not reach statistical significance (Table 3). Both the AUC values for the WBB AP
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
B.-C. Kwok et al. / Clinical Biomechanics xxx (2015) xxx–xxx Table 1 Characteristics of non-fallers (n = 55) and fallers (n = 18). Characteristics
Non-fallers
Age (years) Body mass index (kg m−2) Gender — Female, n (%) Ethnic — Chinese, n (%) Abbreviated Mental Test score (/10) Medical history — co-morbidities, n (%) 0 1 2 ≥3
69.7 (7.8) 22.8 (3.7) 45 (81.8) 54 (98.2) 8.9 (1.1) 6 (10.9) 10 (18.2) 15 (27.2) 24 (43.6)
3
Table 3 Predictive validity of each screening measure for future falls. Fallers
OR (95% CI)a
P value
AUC (95% CI)
0.62 (0.36 to 1.10)
0.10
0.61 (0.46 to 0.77)
1.31 (0.90 to 1.96) 0.51 (0.22 to 1.25)
0.16 0.14
0.64 (0.50 to 0.79) 0.61 (0.46 to 0.77)
0.01
0.67 (0.52 to 0.82)
0.03
0.71 (0.56 to 0.85)
P value
Screening measures
70.7 (5.2) 21.6 (3.7) 17 (94.4) 17 (94.4) 8.7 (1.3)
0.64 0.26 0.20 0.40 0.33
Clinical balance test Short Physical Performance Battery score (/12) Timed-Up-and-Go (s) Gait speed (m s−1)
0 (0) 4 (22.2) 6 (33.3) 8 (44.4)
0.83
Wii Balance Board test (unadjusted) 1.98 (1.16 to 3.40) Anteroposterior velocity (cm s−1) −1 Mediolateral velocity (cm s ) 2.80 (1.10 to 7.13) a
Medications per day, n (%) 0 1 2 ≥3 Vision — Melbourne-edge test (dB) Had a fall in the past 1 year, n (%) Screening measures Short Physical Performance Battery score (/12) Timed-Up-and-Go (s) Gait speed (ms−1) WBB anteroposterior velocity (cm s−1) WBB mediolateral velocity (cm s−1)
14 (25.5) 7 (12.7) 7 (12.7) 27 (49.1) 19.7 (1.6) 28 (51.0)
3 (16.7) 4 (22.2) 2 (11.1) 9 (50.0) 18.9 (2.5) 13 (72.2)
0.61
0.19 0.12
10.6 (1.6)
9.8 (2.0)
0.14
9.10 (4.07) 1.27 (0.44) 1.02 (0.26) 0.73 (0.13)
10.91 (5.47) 1.10 (0.42) 1.27 (0.45) 0.82 (0.14)
0.14 0.14 0.049 0.02
Continuous variables were analyzed with t-test and presented as mean (SD). Categorical variables were analyzed with chi-squared test. WBB: Wii Balance Board.
and ML velocity measures were statistically significant: 0.67 and 0.71 respectively (P b 0.05). At the multivariable level (Table 4), only WBB AP velocity retained its predictive ability of falls (OR 2.21, 95% CI 1.18 to 4.14, P = 0.01), whereas WBB ML velocity did not (OR 2.43, 95% CI 0.86 to 6.90, P = 0.09). 4. Discussion This prospective pilot study identified and established the predictive validity of WBB-derived CoP velocity measures for falls. We found that the WBB balance measure – in particular, the velocity AP measure – could potentially complement the TUG to predict future falls in community-dwelling older adults. 4.1. Relationship between clinical-based and WBB measures Each of the clinical-based measures was moderately correlated to each other but not with the WBB measures (Table 2). The weak relationship between the clinical-based measure and the WBB balance measure showed that the two methods were non-redundant with one another. Our study finding was similar to that of recent literature investigating the relationship between clinical-based and laboratory balance measures (Nguyen et al., 2012). The main difference between these two methods of measures was that the WBB could quantify gross to subtle sway displacements, while clinical-based measures could not identify Table 2 Non-parametric correlation of screening measures.
ORs estimate the odds of future falls at the 75th vs the 25th percentile for all continuous predictors. For example, increasing the mediolateral velocity variable from its lower quartile (0.65 cm s−1) to its higher quartile (0.84 cm s−1) was associated with a 2.8-fold (95%CI, 1.10- to 7.13-fold) increase in the odds of future falls.
by visual inspection (Mancini and Horak, 2010). Thus, in community fall screening, the WBB balance measure could provide valuable information to complement existing clinical-based measures to identify individuals with fall risk. 4.2. Clinical-based measures The clinical-based measures did not predict falls among communitydwelling older adults. Despite a combination of balance, gait speed and lower limb strength measures, the SPPB was not predictive of falls. The TUG measure was also not predictive of falls in our study, similar to previous literature (Boulgarides et al., 2003; Haines et al., 2008), of which one study attributed the null findings to a possible ceiling effect in higher-functioning older adults (Boulgarides et al., 2003). Higherfunctioning older adults are at risk of fall but would not be identified based on cut-off values of 14 s in one study and 15 s in another (Shumway-Cook et al., 2000; Whitney et al., 2005). Finally, as suggested by the Reviewer, another potential explanation for the null results is that because clinical-based measures are multi-construct in nature, their “non-predictive” components may attenuate the overall predictive value of the test. Future studies should examine this possibility. 4.3. WBB balance measure WBB CoP sway velocity quantifies the speed of the movement required to maintain postural stability, and reflects the postural response of the neuromuscular system to somatosensory, visual and vestibular information. Older adults with poorer balance may produce greater muscle co-contraction or stiffening to maintain balance (Carpenter et al., 2001; Ho and Bendrups, 2002), which would result in rapid movements, and hence conceivably higher CoP velocities. Irrespective of AP or ML directions in this study, the WBB velocity measures had better odds in predicting future falls (Table 3) and these findings concur with existing evidence using laboratory force-plate to identify fallers (Bigelow and Berme, 2011; Schneider et al., 2011). However, when adjusted for BMI and in complementing the TUG measure (Table 4), only the AP velocity measure remained predictive of falls. In other words, the predictive effects of AP velocity for future falls were not confounded by those of TUG and BMI; hence, AP velocity could complement the TUG in fall screening. Earlier literature had shown that the AP velocity was higher among older adults with impaired balance (Baloh et al., 1995; Pajala et al., 2008). 4.4. Study implications and limitation
TUG: Timed-Up-and-Go, SPPB: Short Physical Performance Battery, WBB: Wii Balance Board, AP: anteroposterior, ML: mediolateral. ⁎ P b 0.001. ⁎⁎ P b 0.05.
The study provided the first evidence for the WBB-derived CoP velocity measures to predict future falls in a sample of communitydwelling older adults. All participants completed two 30-second standing trials on the WBB without adverse events, supporting the time-
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
4
B.-C. Kwok et al. / Clinical Biomechanics xxx (2015) xxx–xxx
and effort, and the co-authors of previously published literature that led to this study.
Table 4 Logistic regression adjusted for BMI and TUG to predict future falls. Variables
25th percentile
75th percentile
OR (95% CI)a
P value
BMI (kg m−2) TUG (s) AP velocity (cm s−1)
20.2 6.75 0.86
24.6 10.3 1.16
0.51 (0.22 to 1.16) 1.24 (0.81 to 1.91) 2.21 (1.18 to 4.14)
0.11 0.32 0.01
BMI (kg m−2) TUG (s) ML velocity (cm s−1)
20.2 6.75 0.65
24.6 10.3 0.84
0.89 (0.40 to 1.95) 1.27 (0.83 to 1.94) 2.43 (0.86 to 6.90)
0.76 0.28 0.09
References
BMI: body mass index, TUG: Timed-Up-and-Go, AP: anteroposterior, ML: mediolateral. a ORs scaled by the respective screening measure interquartile range.
efficiency and feasibility of WBB testing in field applications. To the clinicians, the WBB screening tool potentially provides a quick and valid measure of balance and fall risk; to the health policy-makers, the WBB may be a technologically advanced yet an affordable screening investment in primary care (WBB and a laptop/tablet combined cost b$500USD) as compared to laboratory-based force plate (more than $10,000USD). Our study has limitations. One study limitation was the lack of home exercise adherence monitoring in the prospective 1-year follow-up, which might have influenced the fall incidence. Another limitation is that during WBB testing, we evaluated bipedal standing so that even older adults with substantial balance limitations could perform the test safely. Nevertheless, it may be informative to evaluate other balance tasks, for example, tandem stances on the WBB, given their abilities to provide greater challenge to balance control. A final limitation of the explorative study was the low sample size. 5. Conclusion The present study supported the predictive validity of the WBB measures for future falls. Specifically, the WBB-derived CoP measures – in particular, the velocity AP measure – could potentially complement the TUG to predict future falls in community-dwelling older adults. Authors' contributions All authors participated in the study design. RAC created the data collection and analysis system. BCK carried out the data collection. YHP and BCK performed the statistical analysis. BCK drafted the manuscript. All authors provided input into the manuscript draft and have read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Acknowledgments The SingHealth Foundation research grant (Grant number: SHF/ FG397S/2009), the Singapore General Hospital research grant (Grant number: SRG04/2010) and the Singapore Physiotherapy Association research grant (Grant number: RF09-005) funded the study. The funders were not involved in the study. We would like to thank Ms Prithvi Balaji (Physiotherapist) for her role as an outcome assessor in the EFFECT study, Mr Kwok Boon Yong (Lecturer) from the School of Communication, Arts and Social Sciences, Singapore Polytechnic for providing language and editorial input, the participants who volunteered their time
Baloh, R.W., Spain, S., Socotch, T.M., Jacobson, K.M., Bell, T., 1995. Posturography and balance problems in older people. J. Am. Geriatr. Soc. 43, 638–644. Bigelow, K.E., Berme, N., 2011. Development of a protocol for improving the clinical utility of posturography as a fall-risk screening tool. J. Gerontol. A Biol. Sci. Med. Sci. 66, 228–233. Botolfsen, P., Helbostad, J.L., Moe-Nilssen, R., Wall, J.C., 2008. Reliability and concurrent validity of the Expanded Timed Up-and-Go test in older people with impaired mobility. Physiother. Res. Int. 13, 94–106. Boulgarides, L.K., McGinty, S.M., Willett, J.A., Barnes, C.W., 2003. Use of clinical and impairment-based tests to predict falls by community-dwelling older adults. Phys. Ther. 83, 328–339. Carpenter, M.G., Frank, J.S., Silcher, C.P., Peysar, G.W., 2001. The influence of postural threat on the control of upright stance. Exp. Brain Res. 138, 210–218. Clark, R.A., Bryant, A.L., Pua, Y., McCrory, P., Bennell, K., Hunt, M., 2010. Validity and reliability of the Nintendo Wii Balance Board for assessment of standing balance. Gait Posture 31, 307–310. Close, J.C., Lord, S.R., 2011. Fall assessment in older people. BMJ 343, d5153. Fleming, J., Matthews, F.E., Brayne, C., 2008. Falls in advanced old age: recalled falls and prospective follow-up of over-90-year-olds in the Cambridge City over-75's Cohort study. BMC Geriatr. 8, 6. Freire, A.N., Guerra, R.O., Alvarado, B., Guralnik, J.M., Zunzunegui, M.V., 2012. Validity and reliability of the short physical performance battery in two diverse older adult populations in Quebec and Brazil. J. Aging Health 24, 863–878. Guralnik, J.M., Simonsick, E.M., Ferrucci, L., et al., 1994. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J. Gerontol. 49, M85–M94. Haines, T., Kuys, S.S., Morrison, G., Clarke, J., Bew, P., 2008. Balance impairment not predictive of falls in geriatric rehabilitation wards. J. Gerontol. A Biol. Sci. Med. Sci. 63, 523–528. Harrell Jr., F.E., 2001. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. Springer, New York. Herala, M., Luukinen, H., Honkanen, R., Koski, K., Laippala, P., Kivela, S.L., 2000. Soft tissue injury resulting from falling predicts a future major falling injury in the home dwelling elderly. J. Epidemiol. Community Health 54, 557–558. Ho, C.Y., Bendrups, A.P., 2002. Ankle reflex stiffness during unperceived perturbation of standing in elderly subjects. J. Gerontol. A Biol. Sci. Med. Sci. 57, B344–B350. Kwok, B.C., Mamun, K., Chandran, M., Wong, C.H., 2011. Evaluation of the Frails' Fall Efficacy by Comparing Treatments (EFFECT) on reducing fall and fear of fall in moderately frail older adults: study protocol for a randomised control trial. Trials 12, 155. Lee, J., Geller, A.I., Strasser, D.C., 2013. Analytical review: focus on fall screening assessments. PM R 5, 609–621. Mancini, M., Horak, F.B., 2010. The relevance of clinical balance assessment tools to differentiate balance deficits. Eur. J. Phys. Rehabil. Med. 46, 239–248. Nguyen, U.S., Kiel, D.P., Li, W., et al., 2012. Correlations of clinical and laboratory measures of balance in older men and women. Arthritis Care Res. 64, 1895–1902. Ottenbacher, K.J., 1998. Quantitative evaluation of multiplicity in epidemiology and public health research. Am. J. Epidemiol. 147, 615–619. Pajala, S., Era, P., Koskenvuo, M., Kaprio, J., Tormakangas, T., Rantanen, T., 2008. Force platform balance measures as predictors of indoor and outdoor falls in communitydwelling women aged 63–76 years. J. Gerontol. A Biol. Sci. Med. Sci. 63, 171–178. Podsiadlo, D., Richardson, S., 1991. The timed “Up & Go”: a test of basic functional mobility for frail elderly persons. J. Am. Geriatr. Soc. 39, 142–148. Power, V., Van de Ven, P., Nelson, J., Clifford, A.M., 2014. Predicting falls in communitydwelling older adults: a systematic review of task performance-based assessment tools. Physiother. Pract. Res. 35, 3–15. Pua, Y.H., Ong, P.H., Lee, A.Y., Tan, J., Bryant, A.L., Clark, R.A., 2013. Preliminary prediction model for fear-induced activity limitation after total knee arthroplasty in people 60 years and older: prospective cohort study. Arch. Phys. Med. Rehabil. 94, 503–509. Schneider, P., Schwab, M., Hanscheid, H., 2011. Identification of factors associated with risk of fall using a force platform and power spectrum analysis technique. J. Biomech. 44, 2008–2012. Sheehan, K.J., O'Connell, M.D., Cunningham, C., Crosby, L., Kenny, R.A., 2013. The relationship between increased body mass index and frailty on falls in community dwelling older adults. BMC Geriatr. 13, 132. Shumway-Cook, A., Brauer, S., Woollacott, M., 2000. Predicting the probability for falls in community-dwelling older adults using the Timed Up & Go Test. Phys. Ther. 80, 896–903. Whitney, J.C., Lord, S.R., Close, J.C., 2005. Streamlining assessment and intervention in a falls clinic using the Timed Up and Go Test and physiological profile assessments. Age Ageing 34, 567–571. Zecevic, A.A., Salmoni, A.W., Speechley, M., Vandervoort, A.A., 2006. Defining a fall and reasons for falling: comparisons among the views of seniors, health care providers, and the research literature. The Gerontologist 46, 367–376.
Please cite this article as: Kwok, B.-C., et al., Novel use of the Wii Balance Board to prospectively predict falls in community-dwelling older adults, Clin. Biomech. (2015), http://dx.doi.org/10.1016/j.clinbiomech.2015.03.006
