Chinese Nursing Research xxx (2016) 1e6

H O S T E D BY Contents lists available at ScienceDirect

Chinese Nursing Research

journal homepage: http://www.journals.elsevier.com/chinese-nursing-research

Original article

Tai Chi as an intervention to reduce falls and improve balance function

in the elderly: A meta-analysis of randomized controlled trials
Yuan Zhao*, Yan Wang
Tianjin University of Traditional Chinese Medicine, Tianjin 300193, China

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To systematically evaluate the effectiveness of fall prevention and balance function in the
Received 29 January 2015 elderly.
Received in revised from Methods: Databases, including PubMed, Web of Science, Cochrane Library, Chinese Biomedical Literature
15 February 2015
Database (CBM), and CNKI were electronically searched, and the relevant references of the included
Accepted 27 October 2015
Available online xxx
papers were also manually searched. Two reviewers independently screened the articles according to the
inclusion and exclusion criteria, extracted the data, and assessed the methodological quality. A meta-
analysis was performed using the Cochrane Collaboration's RevMan 5.1 software.
Keywords:
Tai Chi
Results: Six randomized controlled trials (RCTs) involving 2796 participants were included. The results of
Balance function the meta-analysis showed that compared with a physiotherapy intervention, Tai Chi could significantly
Elderly reduce the incidence rates of falls [relative risk (RR) ¼ 0.82, 95% confidence interval (CI) (0.73, 0.92)],
Fall while there were significant differences in the Timed Up and Go test, Functional Reach Test and Berger
Meta-analysis Balance Scale.
Systematic review Conclusions: Tai Chi is effective in reducing the risk of falls and improving balance in the elderly.
Randomized controlled trial © 2016 Shanxi Medical Periodical Press. Production and hosting by Elsevier B.V. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction to-moderate intensity, Tai Chi improves cardiopulmonary capacity,

muscle strength, postural control, spinal flexibility and balance.
Each year, approximately 30% of community-living adults aged This meta-analysis focused on reviewing and updating the current
65 and older experience a fall. Falls are the second main cause of evidence on using Tai Chi as an intervention for improving balance
accidents with injuries among the elderly, and the medical waste and reducing falls in the elderly.
and economic costs associated with non-lethal falls are consider-
able. In addition, falls among the elderly are considered to be un-
2. Methods
avoidable and an important research subject in public health.1
Some previous studies have indicated that with advancing age,
2.1. Types of studies
balance decreases more sharply than other physiological func-
tions,2 especially among people over the age of 60 years, in whom
Studies were included for review only if they were randomized
the ability to balance is significantly impaired and leads to an
controlled trials (RCTs). We did not plan to include the results from
increased risk of falls; this directly influences the elderly's physical
quasi-RCTs in the analyses, but we may discuss them in the text if
and psychological health.3 Results of many studies have shown that
limited RCT evidence is available.
physical exercise can improve motor function, the ability to bal-
ance, the ability to walk, and the stability and posture of the
elderly.4 As a traditional Chinese sport, Tai Chi exercises consist of a 2.2. Types of participants
series of sequential, graceful, and balanced movements that are
executed in a slow, meditative, and relaxed manner.5 With its mild- According to the definition of the elderly by the World Health
Organization, we included individuals above 60 years of age who
live in nursing institutions or local communities and have not
* Corresponding author.
practiced Tai Chi in the previous 12 months. Participants were
E-mail address: oneinseven1988@163.com (Y. Zhao). excluded if they had a degenerative neurological condition, such as
Peer review under responsibility of Shanxi Medical Periodical Press. Parkinson's disease, dementia, or a severely debilitating stroke;

http://dx.doi.org/10.1016/j.cnre.2015.10.003
2095-7718/© 2016 Shanxi Medical Periodical Press. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article in press as: Zhao Y, Wang Y, Tai Chi as an intervention to reduce falls and improve balance function in the elderly: A meta-
analysis of randomized controlled trials, Chinese Nursing Research (2016), http://dx.doi.org/10.1016/j.cnre.2015.10.003
2 Y. Zhao, Y. Wang / Chinese Nursing Research xxx (2016) 1e6
severe arthritis; or marked vision impairments or if they were
unable to walk across a room independently.
2.3. Types of interventions
2.3.1. TC group
Tai Chi exercises served as the intervention and were adminis-
tered for 3 months or more with instructions of an experienced Tai
Chi instructor. Participants were not given any particular in-
structions about performing Tai Chi outside the class.
2.3.2. Control group
Controls were offered regular exercises (RE) or physical treat-
ments (PT), including resistance training, balance training, strength
training, or walking training Types of outcome measures.
2.4. Outcomes
2.4.1. Primary outcomes
The primary outcome measure was the number of falls.
2.4.2. Secondary outcome
The secondary outcomes included functional balance and
physical performance. Functional balance measures included the
Berg Balance Scale (BBS), which evaluates a person's ability to
perform 14 different tasks of increasing difficulty, and the Function
Reach (FR) measure, which assesses the maximal distance a person
can reach forward beyond an arm's length while maintaining a
fixed base of support in a standing position. The physical perfor-
mance tests involved the Timed Up and Go Test (TUG), which is
reliable and valid in quantifying functional mobility and was used
to assess mobility.
2.5. Search methods for identification of studies
The extensive nature of this topic was reflected in the search of a
wide range of resources, both electronic and non-electronic,
without language restrictions, covering all relevant databases,
including PubMed, Web of Science, Cochrane Library, Chinese
Biomedical Literature Database (CBM), and CNKI. For all of the
databases, we used the highly sensitive search strategy for ran-
domized controlled trials (RCTs), as suggested in the Cochrane
Handbook. All of the relevant articles found were identified on
PubMed using the ‘related articles’ feature, and a further search was
carried out for newly published articles.
The PubMed search strategy included the following search
terms: ‘Aged’(Mesh) OR ‘aged’(Title/Abstract) OR ‘elderly’ (Title/
Abstract) AND [‘Tai Ji’(Mesh) OR ‘Tai Ji’(Title/Abstract) OR ‘Tai
Chi’(Title/Abstract)] AND [‘clinical trial’(Publication Type) OR
‘clinical trials as topic’(MeSH Terms) OR ‘clinical trial’(All Fields) OR
‘random’(All Fields) OR ‘randomized’(All Fields) OR ‘random-
ization’(All Fields) OR ‘randomized’(All Fields) OR ‘randomly’(All
Fields) OR ‘randomness’(All Fields)].
2.6. Data collection and analysis
2.6.1. Selection of studies
All of the titles and/or abstracts generated by the searches were
screened by pairs of authors for potentially relevant studies. The
full-length articles of the selected titles and/or abstracts were
assessed for eligibility. Disagreement was resolved by consensus or
third party adjudication.
Please cite this article in press as: Zhao Y, Wang Y, Tai Chi as an intervention to reduce falls and improve balance function in the elderly: A meta-
analysis of randomized controlled trials, Chinese Nursing Research (2016), http://dx.doi.org/10.1016/j.cnre.2015.10.003
2.6.2. Data extraction and management
Two authors independently extracted the data from eligible
trials using a customized data extraction tool. We recorded
information about the trial design, main characteristics of
the participants, intervention modality assessed, and outcomes.
Disagreement on the data extracted was resolved by consensus or
third party adjudication. We contacted the authors of the studies
where there was inadequate reporting of data to enable clarifica-
tion and, where appropriate, to allow pooling.
2.6.3. Assessment of the risk of bias in the included studies
Two authors independently assessed the risk of bias for each
study using the criteria outlined in the Cochrane Handbook for
Systematic Reviews of Interventions (Higgins 2011). We resolved
any disagreement by involving a third author in the discussion. We
assessed the following domains: (1) Random sequence generation
(checking for possible selection bias); (2) allocation concealment
(checking for possible selection bias); (3) blinding (checking for
possible performance bias); (4) incomplete outcome data (checking
for possible attrition bias due to the amount, nature and handling of
incomplete outcome data); (5) selective reporting (checking for
reporting bias), and (6) other bias (checking for bias due to prob-
lems not covered by areas 1e5).
2.7. Measures of treatment effect
For each trial, the risk ratios and 95% confidence intervals were
calculated for dichotomous outcomes, and the mean differences
(MD) and 95% confidence intervals were calculated for continuous
outcomes (reporting mean and standard deviation or standard error
of the mean). The standardized mean differences (SMD) and 95%
confidence intervals were calculated when combining the results
from studies using different ways of measuring the same concept.
The change in scores was reported separately, as these cannot be
incorporated into a meta-analysis of standardized mean differences.
2.8. Assessment of heterogeneity
Heterogeneity between comparable studies was tested using
visual inspection of the forest plot and a standard chi-square test
and was considered statistically significant at p < 0.1; this was done
with consideration of the value of the I2 statistic, and a value greater
than 50% may indicate substantial heterogeneity.
2.9. Data synthesis
Appropriate statistical analyses were performed using Review
Manager in accordance with the Cochrane Handbook for System-
atic Reviews of Interventions.6 Where available and appropriate,
quantitative data for the outcomes listed in the inclusion criteria
are presented in the analyses. Where appropriate, the results of the
comparable groups of studies were pooled using the fixed-effect
model and the 95% confidence intervals were calculated. In the
presence of substantial heterogeneity, or an I2 statistic greater than
50%, the results of comparable groups of studies were pooled using
the random-effects model and the 95% confidence intervals were
calculated.
2.10. Sensitivity analysis
We planned to conduct sensitivity analyses to explore the effect
of the trial quality as assessed by the concealment of allocation,
high attrition rates, or both, with poor quality studies being
excluded from the analyses to assess whether this made any dif-
ference in the overall result.
 Y. Zhao, Y. Wang / Chinese Nursing Research xxx (2016) 1e6 3

Fig. 1. Flowchart of the procedure for the inclusion of articles for the meta-analysis.

Table 1
Characteristics of the included studies.

Author Area Age (years) (m ± SD) Sample size (n) Intervention Outcome

TC group Control group TC group Control group

Taylor 20127 New Zealand 74 ± 6 220 231 TC PT Fall counts, TUG

Tousignant 20128 Canada 79 ± 6 76 76 TC PT TUG, BBS
Faber 20069 Netherlands Mean age 84.9 80 92 TC RE Fall counts
Wolf 200610 American Mean age 80.9 158 153 TC RE FR
Voukelatos 200711 Australia 69 ± 6.5 353 349 TC RE Fall counts
Li 200512 Poland 77.5 ± 5 125 131 TC PT Fall counts, TUG, FR, BBS

Note: RE: Regular Exercise; PT: Physiotherapy; TUG: Timed Up and Go Test; BBS: Berg Balance Score; FR: Functional Reach.

3. Results for further assessment for their inclusion in the review. Finally, we
included six studies with a total of 2796 participants in the meta-
3.1. Results of the search analysis (Fig. 1).

Originally, we retrieved 587 references from our electronic da-

tabases searches, which resulted in 339 unique references after 3.2. Description of the studies
resolving overlaps between the sources that were searched. We
excluded 326 irrelevant references and selected 133e5,7e16 articles Characteristics of the included studies are given in Table 1.

Fig. 2. Risk of bias graph: a review of the authors' judgments on each risk of bias item presented as percentages across all of the included studies.

Please cite this article in press as: Zhao Y, Wang Y, Tai Chi as an intervention to reduce falls and improve balance function in the elderly: A meta-
analysis of randomized controlled trials, Chinese Nursing Research (2016), http://dx.doi.org/10.1016/j.cnre.2015.10.003
4 Y. Zhao, Y. Wang / Chinese Nursing Research xxx (2016) 1e6
Fig. 3. Risk of bias summary: a review of the authors' judgments on each risk of bias
item for each included study.
3.3. Risk of bias in the included studies
All of the studies stated that the participants were randomly
assigned (e. g., computer-generated table of random numbers, draw
of lots, flip of a coin), and some studies stated that they used
alternate group assignment as an allocation method. For four
studies, the sequence table was held in a sealed envelope in a
locked filing cabinet until study completion. Four studies were
explicitly single-blinded, using a blinded assessment. All of the
studies had low or even no drop-out rates and reported the reasons
for dropouts. Three analyses were conducted on an intention-to-
treat analysis. We did not search for the study protocols to
compare to the study publication; however, for all of the studies,
Fig. 4. Forest plot of comparison: fall counts.
Please cite this article in press as: Zhao Y, Wang Y, Tai Chi as an intervention to reduce falls and improve balance function in the elderly: A meta-
analysis of randomized controlled trials, Chinese Nursing Research (2016), http://dx.doi.org/10.1016/j.cnre.2015.10.003
the outcomes listed in the methods were reported in the results.
For all of the studies, there was a low risk of bias (Figs. 2 and 3).
3.4. Effects of interventions
3.4.1. Falls
Four studies with a total of 1443 participants compared the
incidence of falls, including the number of falls, frequency of falls,
and fallers. The pooled estimate of the favors TC group [relative risk
(RR) ¼ 0.82, 95% confidence interval (CI) (0.73, 0.92)], without ev-
idence of statistical heterogeneity (I2 ¼ 19%). Li et al. found that the
number of falls was reduced in the TC group (38 falls) compared to
the exercise stretching group (73 falls) during the 6-month post-
intervention follow-up. Voukelatos et al. reported a 50% relative
risk (RR) reduction of two or more falls among those who partici-
pated in the TC group compared to the control group (Fig. 4).
3.4.2. Gait
When the primary outcomes were measured immediately post-
intervention, the TC exercise achieved a statistically significant
reduction in the time taken to perform a TUG [SMD ¼ 0.23, 95%CI
(0.09, 0.37), 780 participants, 3 studies]. The fixed effects model
was used to pool the data (c2 ¼ 0.89, I2 ¼ 0%) (Fig. 5).
3.4.3. Balance
We also observed effects that were attributable to the TC group,
showing significantly improved scores for Functional Reach
[SMD ¼ 0.38, 95%CI (0.19, 0.57), 448 participants, 2 studies]
compared with the control group, without evidence of statistical
heterogeneity (I2 ¼ 0%) (Fig. 6).
Additionally, the TC exercise achieved a statistically significant
improvement in the Berg Balance Score at the end of the inter-
vention [MD ¼ 2.45, 95%CI (1.47, 3.43), 345 participants, 2 studies],
without evidence of statistical heterogeneity (I2 ¼ 28%) (Fig. 7).
4. Discussion
This review included 6 RCTs that were from a variety of sour-
ces, had similar clinical settings, and showed promising results for
the use of Tai Chi in preventing falls in the elderly. According to
the integrated and comprehensive evaluation, the results of this
review show that taking pertinent treatment measures can reduce
the possibility of a fall by 30%e40%.17 Wolfson et al.18 conducted a
collection of analyses and divided sports therapy into 6 main
types: walking, balance, and functional exercises; strength
training; resistance training; Tai Chi; normal psychical exercises,
and endurance training. The results showed that sports therapy
had a similar effect across high and low risk groups. However, the
analysis on the results of this research showed that Tai Chi was
able to improve patients' balance and walking capacity by
 Y. Zhao, Y. Wang / Chinese Nursing Research xxx (2016) 1e6 5

Fig. 5. Forest plot of comparison: timed up and go test.

Fig. 6. Forest plot of comparison: functional reach.

Fig. 7. Forest plot of comparison: Berg balance score.

increasing muscle strength and improving the ability to keep one's
balance and enhance one's flexibility as well as by increasing one's
endurance to reduce and prevent the falls; this was statistically
significantly better than the regular exercise or physical therapy
groups.
The results of this study proved that Tai Chi could improve the
ability to stay balanced and increase muscle strength, flexibility and
reacting time, which can help protect the elderly from potential
falls. (1) Tai Chi is a form of exercise that can improve the flexibility
of the body and mind, as well as adjust breathing, and can increase
the cooperation among the hands, eyes and whole body with the
mind. Tai Chi can boost the control of the mind on the body and
improve the reaction capacity of the elderly. (2) Tai Chi exercises
incorporate whole body exercises and some isotonic exercises.
Performing Tai Chi exercises can help build muscles on the back and
the lower limbs, which can subsequently increase muscle strength.
(3) The steps of Tai Chi focus on the exchange between deficiency
and excess, and the so-called ‘deficiency and excess’ (Xu Shi in
Chinese) is the proportion of the body weight excreted on the feet;
the foot that withstands more body weight is referred to as ‘excess,’
and the foot with less body weight is referred to as ‘deficiency.’ ‘The
exchange between deficiency and excess’ actually refers to the
change of body weight between each foot. This theory fully em-
phasizes the capacity to control the distribution of the body weight
and adjust the gestures. With the guidance of this theory and
cooperation of a series of movements made while standing with a
single leg, standing with two legs, or standing with the other leg,
the capacity of the elderly to stay balanced and adjust their steps is
improved. (4) The Tai Chi pattern includes many movements, which
require people to go down on their knees and their waists; this is
good for many joints because it increases the movement of the
joints and improves flexibility.
As in all systematic reviews, this study is susceptible to bias.
Publication bias was likely minimized by the use of a broad search
strategy, but some language bias may still exist. Reliability of the
risk of bias assessment and data extraction was improved because
two researchers performed procedures independently before a
consensus was obtained. Bias due to the 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 heterogeneous sources due to
the small number of studies.
This evaluation still has limitations: first, some parameters are
of greater importance, such as those of progression; with the loss of
these parameters, we were unable to combine and analyze these
parameters; second, the timing of the research varied among the
studies, which may impact the authenticity of the research; and
third, the type of Tai Chi exercises and the forms of physical therapy
that were performed in the studies varied; in addition, the time and
intensity of the exercises were not consistent, which may impact
the authenticity of the meta-analysis results.
Conflicts of interest
All contributing authors declare no conflicts of interest.
References
1. Rose DJ. Preventing falls among older adults: no “one size suits all” interven-
tion strategy. J Rehabil Res Dev. 2008;45:1153e1166.
2. Neyens JC, Dijcks BP, Twisk J, et al. A multifactorial intervention for the pre-
vention of falls in psychogeriatric nursing home patients, a randomised
controlled trial (RCT). Age Ageing. 2009;38:194e199.

Please cite this article in press as: Zhao Y, Wang Y, Tai Chi as an intervention to reduce falls and improve balance function in the elderly: A meta-
analysis of randomized controlled trials, Chinese Nursing Research (2016), http://dx.doi.org/10.1016/j.cnre.2015.10.003
6 Y. Zhao, Y. Wang / Chinese Nursing Research xxx (2016) 1e6

3. Greenspan AI, Wolf SL, Kelley ME, O'Grady M. Tai chi and perceived health
status in older adults who are transitionally frail: a randomized controlled trial.
Phys Ther. 2007;87:525e535.
4. Frye B, Scheinthal S, Kemarskaya T, Pruchno R. Tai Chi and low impact exercise:
effects on the physical functioning and psychological well-being of older
people. J Appl Gerontol. 2007;26:433e453.
5. Chyu MC, James CR, Sawyer SF, et al. Effects of tai chi exercise on posturog-
raphy, gait, physical function and quality of life in postmenopausal women
with osteopaenia: a randomized clinical study. Clin Rehabil. 2010;24:
1080e1090.
6. Higgins PT, Green S. Cochrane Handbook for Systematic Reviews of Interventions
Version 5.1.0 Updated March 2011. The Cochrane Collaboration; 2011. www.
cochrane-handbook.org, 2012.12.28.
7. Taylor D, Hale L, Schluter P, et al. Effectiveness of tai chi as a community-based
falls prevention intervention: a randomized controlled trial. J Am Geriatr Soc.
2012;5:841e848.
8. Tousignant M, Corriveau H, Roy PM, et al. The effect of supervised Tai Chi
intervention compared to a physiotherapy program on fall-related clinical
outcomes: a randomized clinical trial. Disabil Rehabil. 2012;34:196e201.
9. Faber MJ, Bosscher RJ, Chin A, Paw MJ, van Wieringen PC. Effects of exercise
programs on falls and mobility in frail and pre-frail older adults: a multicenter
randomized controlled trial. Arch Phys Med Rehabil. 2006;7:885e896.
10. Wolf SL, O'Grady M, Easley KA, Guo Y, Kressig RW, Kutner M. 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. 2006;61:
184e189.
11. Voukelatos A, Cumming RG, Lord SR, Rissel C. A randomized, controlled trial of
tai chi for the prevention of falls: the central Sydney Tai Chi trial. J Am Geriatr
Soc. 2007;8:1185e1191.
12. Li F, Harmer P, Fisher KJ, et al. Tai Chi and fall reductions in older adults: a
randomized controlled trial. J Gerontol A Biol Sci Med Sci. 2005;60:187e194.
13. Lelard T, Doutrellot PL, David P, Ahmaidi S. Effects of a 12-week Tai Chi Chuan
program versus a balance training program on postural control and walking
ability in older people. Arch Phys Med Rehabil. 2010;91:9e14.
14. Woo J, Hong A, Lau E, Lynn H. A randomised controlled trial of Tai Chi and
resistance exercise on bone health, muscle strength and balance in
community-living elderly people. Age Ageing. 2007;36:262e268.
15. Maciaszek J, Osin  ski W, Szeklicki R, Stemplewski R. Effect of Tai Chi on body
balance: randomized controlled trial in men with osteopenia or osteoporosis.
Am J Chin Med. 2007;35:1e9.
16. Li F, Harmer P, Fisher KJ, McAuley E. Tai Chi: improving functional balance and
predicting subsequent falls in older persons. Med Sci Sports Exerc. 2004;36:
2046e2052.
17. Chang JT, Morton SC, Rubenstein LZ, et al. Interventions for the prevention of
falls in older adults: systematic review and meta-analysis of randomised
clinical trials. BMJ. 2004;328:680.
18. Wolfson L, Whipple R, Derby C, et al. Balance and strength training in older adults:
intervention gains and Tai Chi maintenance. J Am Geriatr Soc. 1996;44:498e506.

Please cite this article in press as: Zhao Y, Wang Y, Tai Chi as an intervention to reduce falls and improve balance function in the elderly: A meta-
analysis of randomized controlled trials, Chinese Nursing Research (2016), http://dx.doi.org/10.1016/j.cnre.2015.10.003