The effects of Tai Chi on fall prevention, fear of falling and balance in older people: A meta-analysis

Results Nine trials (representing 2203 participants) were included in the analyses. Compared with exercise controls, TC participants showed significant improvements in fall rates (2 trials included, IRR: 0.51, 95% CI 0.38–0.68) and static balance (2 trials included, SMD: 0.47, 95% CI 0.23–0.72). Compared with non-exercise controls, no improvement was found for TC participants in fall rates (5 trials, IRR: 0.79, 95% CI 0.60–1.03) or static balance (2 trials, SMD: 0.30, 95% CI − 0.50–1.10), but a significant improvement was found for fear of falling (SMD: 0.37, 95% CI = 0.03–0.70). Conclusions Currently there is insufficient evidence to conclude whether TC is effective in fall prevention, decreasing fear of falling and improving balance in people over age 50 years. Keywords Tai Chi Fall prevention Meta-analysis Introduction Each year approximately 30% of community-living persons aged 65 years and older experiences a fall. Of these falls, 55–70% result in physical injury of which 20% require medical attention ( Gillespie et al., 2009; Nevitt et al., 1991; Rubenstein, 2006 ). Estimated costs of falls (for the year 2000) range from almost one billion pounds in the UK to 19.2 billion dollars in the USA ( Scuffham et al., 2003; Stevens et al., 2006 ). Given the individual burden of falls and the societal costs, prevention of falls is desirable. Until now, the merits of Tai Chi (TC) as a fall prevention intervention are promising but remain inconclusive ( Chang et al., 2004; Gillespie et al., 2009; Harmer and Li, 2008; Hong and Li, 2007; Komagata and Newton, 2003; Li et al., 2004b, 2005b; Low et al., 2009; Sherrington et al., 2008; Sjosten et al., 2008; Verhagen et al., 2004; Voukelatos et al., 2007; Wayne et al., 2004; Wu, 2002 ). The integrated physical and cognitive components in TC could represent the additional value of TC compared with other exercise programs which mainly focus on physical aspects only. Besides the improvement of balance, TC might increase ‘self-awareness of balance’ and thereby decrease the fear of falling ( Chan and Barlett, 2000; Howe et al., 2007; Li et al., 2005a; Zijlstra et al., 2007; Zwick et al., 2000 ). Based on the type of TC, exercise dose, intervention duration and the study population, the effects of TC might vary. If TC is to be used as a community-based fall prevention intervention, insight is needed into the effectiveness of TC with regard to different populations, optimal exercise dose and effect maintenance. This meta-analysis assesses the effectiveness of TC as a single intervention on fall rate, fear of falling and balance in healthy older people, to provide insight into the effectiveness of TC regarding setting, intervention dose and duration of follow-up. Methods Search strategy A broad literature search on TC interventions was made covering Medline, Cinahl, Psychlit and the Cochrane Database for Systematic Reviews (CDSR) until January 31 2009. For all databases we used the highly sensitive search strategy for randomized clinical trials (RCTs) as suggested in the Cochrane Handbook. References in relevant reviews and identified RCTs were also screened. Study selection Studies were included when 1) the design was an RCT, 2) participants were aged 50 years or older, 3) the population comprised healthy participants, 4) one of the interventions was a form of TC, and 5) the article included one of the outcome measures of interest (falls, fear of falling, and/or balance). After selection based on title and abstract, the full-text articles were retrieved and two reviewers (AV, IL) independently performed study selection. Studies in English, French, German or Dutch were eligible. Study selection was based on consensus, otherwise a third reviewer (BK) solved any disagreements. Risk of bias assessment Two reviewers (MF, EvR) independently performed methodological assessment of the included trials using the Delphi criteria list ( Verhagen et al., 1998 ) ( Table 1 ). Differences in scores were resolved by consensus or in conjunction with a third party (IL). For one trial in the review, in which two of the current reviewers were both co-authors, two other reviewers performed the methodological assessment ( Logghe et al., 2009 ). All items have a ‘yes’/‘no’/‘don't know’ answer option. If bias was unlikely the criterion was rated positive (‘yes’). In case information was lacking or insufficient the criterion was rated ‘don't know’, and if bias was likely to be negative (‘no’). All ‘yes’ scores were summed to an overall risk of bias score, with higher scores indicating lower risk of bias. The cut-off point between high and low risk was set at 5 points, i.e. ≤ 4 points indicating a high risk of bias and ≥ 5 points indicating a low risk of bias. Data extraction Data on characteristics of the study population, the intervention and outcome measures, i.e. number of falls, estimates of fall rate ratio and mean values (± standard deviations; SD) of fear of falling and balance, were extracted using standardized forms. We refer to ‘static balance’ when balance was measured when the body has a constant (or static) basis of support and we refer to ‘dynamic balance’ when balance was measured during movement. One reviewer (IL) extracted the data; these data were subsequently checked by a second (reviewer AR). Wherever possible, authors of included articles were contacted to complete missing data. Statistical analysis When we considered studies clinically homogeneous regarding intervention, the study population and outcome measures results were combined using random effect models. Where possible, longer follow-up times were used (e.g. 12 months rather than 6 months). The I 2 statistics were calculated, which describe the heterogeneity among study results for each effect. Values of 30–60%, 50–90% and 75–100% represent moderate, substantial and considerable heterogeneity ( Deeks et al., 2008 ). TC versus non-exercise controls and versus exercise controls (e.g. low intensity) were compared separately. For falls the estimates of incidence rate ratios (IRR) were calculated using data on the total number of falls, or the number of falls per person and exposure times. The Hedges' adjusted g effect sizes (ES) were calculated to evaluate the effects of TC on fear of falling and balance. An effect size of 0.2–0.5 was defined as small, 0.5–0.8 as medium and 0.8 and higher was defined as a large effect size ( Schünemann et al., 2008 ). The subgroups of interest were: setting (community dwelling, facility), intervention dose (< 40 sessions, ≥ 40 sessions) and duration of follow-up (post-treatment, follow-up). A greater relative effect of exercise on fall rates was seen in programs that included a higher dose (> 50 hours); 40 sessions is roughly equivalent to 50–60 hours ( Sherrington et al., 2008 ). The software package STATA (version 10.0) was used for the analyses (StataCorp, College Station, Texas, USA) together with Review Manager 5.0. Results Search strategy The search strategy resulted in 524 titles of potentially eligible studies. Based on title and abstract we selected 41 references of which full-text articles were obtained. One full-text article could not be retrieved ( Zhao et al., 2007 ). Finally, 21 papers were included in this systematic review, representing 15 trials ( Audette et al., 2006; Choi et al., 2005; Frye et al., 2007; Li et al., 2004a,b, 2005a,b; Li et al., 2008; Logghe et al., 2009; Nnodim et al., 2006; Nowalk et al., 2001; Sattin et al., 2005; Voukelatos et al., 2007; Wallsten et al., 2006; Wolf et al., 1996, 1997, 2003a,b, 2006; Woo et al., 2007; Zhang et al., 2006 ) ( Fig. 1 ). The results of the Atlanta FICSIT trial were reported in three papers ( Wolf et al., 1996, 1997, 2003a ), as were the results of another trial conducted by Wolf et al. ( Sattin et al., 2005; Wolf et al., 2003b, 2006 ). Li et al. conducted two different trials ( Li, et al., 2004a,b ) of which the results of one trial were reported in three papers ( Li et al., 2004b, 2005a,b ). Risk of bias assessment Disagreement mainly occurred because of reading errors and differences in interpretation. The agreement between the two authors was acceptable (kappa = 0.69). After the consensus meeting no disagreements persisted. The overall score ranged from 3 to 7(median = 6). Nine trials met our definition of low risk of bias ( Table 2 ). In 7 trials the outcome assessor was blinded. None of the trials blinded the care provider or participant; in one trial these aspects remained unclear. Study characteristics The supplementary material presents the characteristics of the included trials. All papers were published between May 1996 and January 2009, and most trials were conducted in the USA (i.e. 10 of 21). Study population In total 2708 older people were included in the present review. Most of the participants were community living (71%, n = 1927). The number of participants ranged from 8 to 353 per study group. The drop-out rates ranged from 2 to 55%. Interventions Three trials included two exercise groups and one (non-exercise) control group. Of the 12 two-arm trials, seven used a non-exercise control intervention and five used an exercise control group. In most trials the participants in the non-exercise control groups were advised to continue their current level of activity and not to practice TC. The dose of TC exercise programs ranged from 16 to 120 hours. Outcome measures Six trials explicitly reported on fall incidents ( Choi et al., 2005; Li, et al., 2004b; Logghe et al., 2009; Nowalk et al., 2001; Voukelatos et al., 2007; Wolf et al., 2003b ), seven trials explicitly gathered data on balance ( Choi et al., 2005; Frye et al., 2007; Li et al., 2008; Nnodim et al., 2006; Voukelatos et al., 2007; Wallsten et al., 2006; Zhang et al., 2006 ) and four trials explicitly on fear of falling ( Choi et al., 2005; Wallsten et al., 2006; Wolf et al., 1996; Zhang et al., 2006 ). Most trials in a community setting measured falls with fall calendars ( Li et al., 2004b; Logghe et al., 2009; Voukelatos et al., 2007; Wolf et al., 1996 ). In a facility setting, one trial used fall calendars ( Wolf et al., 2003b ), one trial monitored weekly fall episodes, and a third registered self-reported or witness-reported falls. Most trials assessed fear of falling with the Falls Efficacy Scale ( Choi et al., 2005; Logghe et al., 2009; Wolf et al., 1996, 2003b; Zhang et al., 2006 ). Eleven trials measured static balance with direct methods (e.g. a force platform) ( Voukelatos et al., 2007; Wolf et al., 1997; Woo et al., 2007 ) or indirect methods (e.g. single-leg stance) ( Audette et al., 2006; Choi et al., 2005; Li, et al., 2004a, 2005b; Li et al., 2008; Nnodim et al., 2006; Wolf et al., 2003b; Woo et al., 2007; Zhang et al., 2006 ). Six trials measured dynamic balance indirectly with quantifiable measures (e.g. Up and Go test) ( Frye et al., 2007; Li, et al., 2005b; Nnodim et al., 2006; Voukelatos et al., 2007; Wolf et al., 2003b ) or observational measures (e.g. Berg Balance Test) ( Logghe et al., 2009 ). Effectiveness of TC versus non-exercise controls Falls Five studies were included in this meta-analysis ( Logghe et al., 2009; Voukelatos et al., 2007; Wolf et al., 1996, 2003b; Woo et al., 2007 ). The pooled estimate of the IRR was 0.79 [95% confidence interval (CI); 0.60–1.03] in favour of TC. There was a substantial level of heterogeneity (I 2 68.6%). No changes in effects were found in the subgroup analyses according to setting, intervention dose and duration of follow-up ( Table 3 ). Fear of falling This meta-analysis was performed with three studies ( Logghe et al., 2009; Wolf et al., 2003b; Zhang et al., 2006 ). A small but significant positive effect of TC was found for fear of falling [ES: 0.37 (95% CI; 0.03–0.70)]. The effect became non-significant in all subgroups, except in the subgroup high intervention dose where the effect increased and remained significant ( Table 3 ). Balance Four studies could be included in the meta-analysis. Two studies provided post-treatment data with direct measurement methods ( Voukelatos et al., 2007; Wolf et al., 1997 ) and two with indirect methods ( Li et al., 2008; Zhang et al., 2006 ). Non-significant effects of TC were found concerning static balance measured with both methods ( Table 3 ). The results of the trials on dynamic balance could not be pooled due to clinical heterogeneous outcome measures. Two studies found a significant improvement in the TC group ( Frye et al., 2007; Voukelatos et al., 2007 ) whereas another study found no effect in the TC group ( Logghe et al., 2009 ). Therefore, the evidence for the effects of TC on dynamic balance is inconclusive. Effectiveness of TC versus exercise controls Falls Our meta-analysis of two studies ( Li, et al., 2004b; Woo et al., 2007 ) showed a pooled estimate of the IRR of 0.51 (95% CI; 0.38–0.68) ( Table 4 ). Fear of falling One study measured fear of falling and the TC participants had significantly lower levels of fear of falling post-treatment; however, this effect disappeared during follow-up ( Li, et al., 2004b ). Balance Our meta-analysis of two studies ( Li, et al., 2004a,b ) showed a small but significant effect on static balance measured with the Single-Leg Stance (ES: 0.47, 95%CI; 0.23–0.72) ( Table 4 ). The data of three studies that measured dynamic balance could not be pooled due to clinical heterogeneity of outcome and non-reported data. One study found a significant improvement ( Li, et al., 2005b ) and two studies found no effect ( Frye et al., 2007; Nnodim et al., 2006 ). Thus, the evidence for the effects of TC on dynamic balance is inconclusive. Discussion Until now no pooled estimates are available for the effects of TC alone on fall reduction. In the present study we found a non-significant overall fall reduction of 21% for TC compared to non-exercise controls. The amount of fall reduction is in line with earlier meta-analyses of exercise interventions on fall reduction, in which a significant 17% reduction in fall was found ( Province et al., 1995; Sherrington et al., 2008 ). We also found a larger effect in a community setting and in exercise interventions using higher doses ( Harmer and Li, 2008 Sherrington et al., 2008 ). Patient characteristics of community-dwelling elderly differed in aspects of activity level, fall risk and mean age, and the intervention characteristics varied with regard to dose and TC styles. For example, the level of activity might be associated with the number of falls, and a higher level of physical activity might reduce the risk of recurrent falling ( Peeters et al., 2009; Wijlhuizen et al., 2008 ). All these aspects may have confounded the effects found in a community setting. We expected to find no differences in effect between TC compared to exercise controls, because the difference between the intervention and the control groups is rather small. Nevertheless, we found a significant fall reduction of 49% in favour of TC. One explanation for this may be the nature of the interventions for the exercise controls. In one study this consisted of low-intensity exercise training without any strength or balance benefits, and in another study the intervention consisted of resistance training ( Li, et al., 2005b; Woo et al., 2007 ). Given the more positive effects on fall reduction found in exercise interventions that challenge balance, it is plausible that the exercise control groups were comparable with placebo controls. Despite this, we cannot explain the large reduction in fall risk of TC versus exercise controls compared with TC versus non-exercise controls. The difference might suggest that exercise control interventions increased fall risk, which seems implausible. Another explanation might be a positive dose–effect relation. Studies evaluating TC versus exercises included only TC interventions with long-term exercises, resulting in a higher dose compared to the dose in the studies on TC versus non-exercise controls. Finally, the unexpected result may have occurred by chance, as only two studies were included in the comparison between TC and exercise controls. Although a fall reduction of 21% seems clinically relevant, it remains unclear whether a population-based implementation of a TC group intervention will be cost effective. Although TC is not an expensive fall prevention intervention, our results suggest that a higher dose of TC might be needed to be effective. Our general findings regarding the positive effects of TC on fear of falling compared with non-exercise controls are in line with earlier reviews ( Sjosten et al., 2008; Zijlstra et al., 2007 ). Also, positive effects become larger with higher intervention doses. However, the two studies that measured falls and fear of falling, found positive effects on fear of falling but not on fall risk. The nature of the relationship between fear of falling and falls remains unclear. Due to the limited number of studies we are unable to draw firm conclusions from our subgroup analysis. No evidence was found for the effect of TC in improving balance compared with non-exercise controls under static conditions. These findings are in line with the Cochrane review of Howe et al. (2007) in which TC was part of an exercise program and no significant improvement in balance was observed. We found a small effect when TC was compared with exercise controls. This finding supports our earlier explanation for the unexpected results on fall reduction in the studies comparing TC with an exercise control group, which seems a good placebo control for balance. However, it seems unlikely that this small positive effect on static balance is responsible for the high fall reduction. In general, evidence for the effect of TC on balance remains unclear. Strength and limitations This is the first study in which data on the effects of TC on falls, fear of falling and balance were pooled and effect estimates provided. With additional subgroup analyses the present study also aimed to gain more insight into the effectiveness of TC with regard to different study populations, intervention characteristics, and effect maintenance. As in all systematic reviews, this study is susceptible to bias. Publication bias was probably minimized by using a broad search strategy, but some language bias may exist. Reliability of the risk of bias assessment and data extraction was improved because two researchers performed procedures independently before consensus was obtained. Bias due to study design and quality was minimized by including only RCTs and by giving an overview of the risk of bias assessment. In the present review, there was little possibility to study the influence of sources of heterogeneity due to the small number of studies. Furthermore, not all studies provided sufficient information about important study characteristics, which hampered subgroup analysis. We believe that these analyses are needed, because the effectiveness of TC might change when using a different intervention dose and/or dose duration and/or another control group. Finally, due to the small number of included studies and the diversity in outcome measures used we could not investigate the relationships between falls, fear of falling and balance directly in meta-regressions. Conclusions Currently, there is insufficient evidence to conclude whether TC is effective in fall prevention, decreasing fear of falling and improving balance in people over age 50 years. However, the presence of a positive dose–effect relation in TC is highly likely. Future research should focus on the role of patient characteristics (e.g. living setting, activity level), intervention dose and effect maintenance on the measured outcomes. Conflict of interest statement The authors declare that there are no conflicts of interest. Sponsors role None. Acknowledgments The authors thank Yvonne van Leeuwen and Pim Luijsterburg (Dept. of General Practice) who performed the methodology assessment of one article, and Sten Willemsen (Dept. of General Practice) for his valuable assistance with the statistical analyses. We also thank the authors that we contacted for supplying us with additional information about the trials they had conducted. Appendix A Supplementary data For the included studies: Number and general characteristics of participants, information on the interventions, follow-up times, outcome measures, number of drop-outs, loss to follow-up, adherence and results regarding falls, fear of falling and balance (Rotterdam, the Netherlands, 2009). Appendix A Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ypmed.2010.06.003 . References Audette et al., 2006 J.F. Audette Y.S. Jin R. Newcomer L. Stein G. Duncan W.R. Frontera Tai Chi versus brisk walking in elderly women Age Ageing 35 2006 388 393 Chan and Barlett, 2000 W.W. Chan D.J. Barlett Effectiveness of Tai Chi as a therapeutic exercise in improving balance and postural control Phys. Occupat. Ther. Geriatr. 17 2000 1 22 Chang et al., 2004 J.T. Chang S.C. Morton L.Z. Rubenstein W.A. Mojica M. Maglione M.J. Suttorp E.A. Roth P.G. Shekelle Interventions for the prevention of falls in older adults: systematic review and meta-analysis of randomised clinical trials BMJ 328 7441 2004 680 Choi et al., 2005 J.H. Choi J.S. Moon R. Song Effects of Sun-style Tai Chi exercise on physical fitness and fall prevention in fall-prone older adults J. Adv. Nurs. 51 2005 150 157 Deeks et al., 2008 J. Deeks J. Higgins D. Altman Chapter 9: Analysing data and undertaking meta-analyses J. Higgins S. Green Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.1 (updated September 2008) The Cochrane Collaboration 2008 Frye et al., 2007 B. Frye S. Scheinthal T. Tatyana Kemarskaya R. Pruchno Tai Chi and low impact exercise: effects on the physical functioning and psychological well-being of older people J. Appl. Gerontol. 26 2007 433 453 Gillespie et al., 2009 L.D. Gillespie M.C. Robertson W.J. Gillespie S.E. Lamb S. Gates R.G. Cumming B.H. Rowe Interventions for preventing falls in older people living in the community Cochrane Database Syst. Rev. 2 2009 CD007146 Harmer and Li, 2008 P.A. Harmer F. Li Tai Chi and falls prevention in older people Med. Sport Sci. 52 2008 124 134 Hong and Li, 2007 Y. Hong J.X. Li Biomechanics of Tai Chi: a review Sports Biomech. 6 2007 453 464 Howe et al., 2007 T.E. Howe L. Rochester A. Jackson P.M. Banks V.A. Blair Exercise for improving balance in older people Cochrane Database Syst. Rev. 4 2007 CD004963 Komagata and Newton, 2003 S. Komagata R. Newton The effectiveness of Tai Chi on improving balance in older adults: an evidence-based review J. Geriatr. Phys. Ther. 26 2003 9 16 Li et al., 2004a F. Li K.J. Fisher P. Harmer D. Irbe R.G. Tearse C. Weimer Tai Chi and self-rated quality of sleep and daytime sleepiness in older adults: a randomized controlled trial J. Am. Geriatr. Soc. 52 2004 892 900 Li et al., 2004b F. Li P. Harmer K.J. Fisher E. McAuley Tai Chi: improving functional balance and predicting subsequent falls in older persons Med. Sci. Sports Exerc. 36 2004 2046 2052 Li et al., 2005a F. Li K.J. Fisher P. Harmer E. McAuley Falls self-efficacy as a mediator of fear of falling in an exercise intervention for older adults J. Gerontol. B Psychol. Sci. Soc. Sci. 60 2005 34 40 Li et al., 2005b F. Li P. Harmer K.J. Fisher E. McAuley N. Chaumeton E. Eckstrom N.L. Wilson Tai Chi and fall reductions in older adults: a randomized controlled trial J. Gerontol. A Biol. Sci. Med. Sci. 60 2005 187 194 Li et al., 2008 J.X. Li D.Q. Xu Y. Hong Effects of 16-week Tai Chi intervention on postural stability and proprioception of knee and ankle in older people Age Ageing 37 2008 575 578 Logghe et al., 2009 I.H. Logghe P.E. Zeeuwe A.P. Verhagen R.M. Wijnen-Sponselee S.P. Willemsen S.M. Bierma-Zeinstra E. van Rossum M.J. Faber B.W. Koes Lack of effect of Tai Chi Chuan in preventing falls in elderly people living at home: a randomized clinical trial J. Am. Geriatr. Soc. 57 2009 70 75 Low et al., 2009 S. Low L.W. Ang K.S. Goh S.K. Chew A systematic review of the effectiveness of Tai Chi on fall reduction among the elderly Arch. Gerontol. Geriatr. 48 2009 325 331 Nevitt et al., 1991 M.C. Nevitt S.R. Cummings E.S. Hudes Risk factors for injurious falls: a prospective study J. Gerontol. 46 1991 M164 M170 Nnodim et al., 2006 J.O. Nnodim D. Strasburg M. Nabozny L. Nyquist A. Galecki S. Chen N.B. Alexander Dynamic balance and stepping versus Tai Chi training to improve balance and stepping in at-risk older adults J. Am. Geriatr. Soc. 54 2006 1825 1831 Nowalk et al., 2001 M.P. Nowalk J.M. Prendergast C.M. Bayles F.J. D'Amico G.C. Colvin A randomized trial of exercise programs among older individuals living in two long-term care facilities: the FallsFREE program J. Am. Geriatr. Soc. 49 2001 859 865 Peeters et al., 2009 G.M. Peeters N.M. van Schoor S.M. Pluijm D.J. Deeg P. Lips Is there a U-shaped association between physical activity and falling in older persons? Osteoporos. Int 2009 10.1007/s00198-009-1053-4 Province et al., 1995 M.A. Province E.C. Hadley M.C. Hornbrook L.A. Lipsitz J.P. Miller C.D. Mulrow M.G. Ory R.W. Sattin M.E. Tinetti S.L. Wolf The effects of exercise on falls in elderly patientsA preplanned meta-analysis of the FICSIT Trials. Frailty and injuries: cooperative studies of intervention techniques Jama 273 1995 1341 1347 Rubenstein, 2006 L.Z. Rubenstein Falls in older people: epidemiology, risk factors and strategies for prevention Age Ageing Suppl 2 2006 ii37 ii41 Sattin et al., 2005 R.W. Sattin K.A. Easley S.L. Wolf Y. Chen M.H. Kutner Reduction in fear of falling through intense Tai Chi exercise training in older, transitionally frail adults J. Am. Geriatr. Soc. 53 2005 1168 1178 Schünemann et al., 2008 H.J.O.A. Schünemann G.E. Vist J.P.T. Higgins J.J. Deeks P. Glasziou G.H. Guyatt Chapter 12: Interpreting results and drawing conclusions J.P.T. Higgins Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.1 (updated Sept 2008) The Cochrane Collaboration 2008 Scuffham et al., 2003 P. Scuffham S. Chaplin R. Legood Incidence and costs of unintentional falls in older people in the United Kingdom J. Epidemiol. Commun. Health 57 2003 740 744 Sherrington et al., 2008 C. Sherrington J.C. Whitney S.R. Lord R.D. Herbert R.G. Cumming J.C. Close Effective exercise for the prevention of falls: a systematic review and meta-analysis J. Am. Geriatr. Soc. 56 2008 2234 2243 Sjosten et al., 2008 N. Sjosten S. Vaapio S.L. Kivela The effects of fall prevention trials on depressive symptoms and fear of falling among the aged: a systematic review Aging Ment. Health 12 2008 30 46 Stevens et al., 2006 J.A. Stevens P.S. Corso E.A. Finkelstein T.R. Miller The costs of fatal and non-fatal falls among older adults Inj. Prev. 12 2006 290 295 Verhagen et al., 1998 A.P. Verhagen H.C. de Vet R.A. de Bie A.G. Kessels M. Boers L.M. Bouter P.G. Knipschild The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus J. Clin. Epidemiol. 51 1998 1235 1241 Verhagen et al., 2004 A.P. Verhagen M. Immink A. van der Meulen S.M. Bierma-Zeinstra The efficacy of Tai Chi Chuan in older adults: a systematic review Fam. Pract. 21 2004 107 113 Voukelatos et al., 2007 A. Voukelatos R.G. Cumming S.R. Lord C. Rissel A randomized, controlled trial of Tai Chi for the prevention of falls: the Central Sydney Tai Chi trial J. Am. Geriatr. Soc. 55 2007 1185 1191 Wallsten et al., 2006 S.M. Wallsten K. Bintrim D.W. Denman J.M. Parrish G. Hughes The effect of Tai Chi chuan on confidence and lower extremity strength and balance in residents living independently at a continuing care retirement community J. Appl. Gerontol. 25 2006 82 95 Wayne et al., 2004 P.M. Wayne D.E. Krebs S.L. Wolf K.M. Gill-Body D.M. Scarborough C.A. McGibbon T.J. Kaptchuk S.W. Parker Can Tai Chi improve vestibulopathic postural control? Arch. Phys. Med. Rehabil. 85 2004 142 152 Wijlhuizen et al., 2008 G.J. Wijlhuizen A.M. Chorus M. Hopman-Rock The 24-h distribution of falls and person-hours of physical activity in the home are strongly associated among community-dwelling older persons Prev. Med. 46 2008 605 608 Wolf et al., 1996 S.L. Wolf H.X. Barnhart N.G. Kutner E. McNeely C. Coogler T. Xu Reducing frailty and falls in older persons: an investigation of Tai Chi and computerized balance training. Atlanta FICSIT Group. Frailty and injuries: cooperative studies of intervention techniques J. Am. Geriatr. Soc. 44 1996 489 497 Wolf et al., 1997 S.L. Wolf H.X. Barnhart G.L. Ellison C.E. Coogler The effect of Tai Chi Quan and computerized balance training on postural stability in older subjects. Atlanta FICSIT Group. Frailty and injuries: cooperative studies on intervention techniques Phys. Ther. 77 1997 371 384 Wolf et al., 2003a S.L. Wolf H.X. Barnhart N.G. Kutner E. McNeely C. Coogler T. Xu F.G. Atlanta Selected as the best paper in the 1990s: reducing frailty and falls in older persons: an investigation of Tai Chi and computerized balance training J. Am. Geriatr. Soc. 51 2003 1794 1803 Wolf et al., 2003b S.L. Wolf R.W. Sattin M. Kutner M. O'Grady A.I. Greenspan R.J. Gregor Intense Tai Chi exercise training and fall occurrences in older, transitionally frail adults: a randomized, controlled trial J. Am. Geriatr. Soc. 51 2003 1693 1701 Wolf et al., 2006 S.L. Wolf M. O'Grady K.A. Easley Y. Guo R.W. Kressig M. Kutner The influence of intense Tai Chi training on physical performance and hemodynamic outcomes in transitionally frail, older adults J. Gerontol. A Biol. Sci. Med. Sci. 61 2006 184 189 Woo et al., 2007 J. Woo A. Hong E. Lau H. Lynn A randomised controlled trial of Tai Chi and resistance exercise on bone health, muscle strength and balance in community-living elderly people Age Ageing 36 2007 262 268 Wu, 2002 G. Wu Evaluation of the effectiveness of Tai Chi for improving balance and preventing falls in the older population—a review J. Am. Geriatr. Soc. 50 2002 746 754 Zhang et al., 2006 J.G. Zhang K. Ishikawa-Takata H. Yamazaki T. Morita T. Ohta The effects of Tai Chi Chuan on physiological function and fear of falling in the less robust elderly: an intervention study for preventing falls Arch. Gerontol. Geriatr. 42 2006 107 116 Zhao et al., 2007 J. Zhao L. Zhang Y. Tian Effect of 6 months of Tai Chi Chuan and calcium supplementation on bone health in females aged 50–59 years J. Exerc. Sci. Fitness 5 2007 88 94 Zijlstra et al., 2007 G.A. Zijlstra J.C. van Haastregt E. van Rossum J.T. van Eijk L. Yardley G.I. Kempen Interventions to reduce fear of falling in community-living older people: a systematic review J. Am. Geriatr. Soc. 55 2007 603 615 Zwick et al., 2000 D. Zwick A. Rochelle A. Choksi J. Domowicz Evaluation and treatment of balance in the elderly: a review of the efficacy of the Berg Balance Test and Tai Chi Quan NeuroRehabil. 15 2000 49 56