Sleep Medicine 33 (2017) 70e75

Contents lists available at ScienceDirect

Sleep Medicine
journal homepage: www.elsevier.com/locate/sleep

Original Article

Effect of Tai Ji Quan training on self-reported sleep quality in elderly

Chinese women with knee osteoarthritis: a randomized controlled
trail
Jiaojiao Lü, Lingyan Huang, Xie Wu, Weijie Fu, Yu Liu*
Key Laboratory of Exercise and Health Sciences of Ministry of Education, Shanghai University of Sport, Shanghai, 200438, China

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

Article history: Objective: The purpose of this study was to explore the effects of a 24-week Tai Ji Quan training program
Received 17 January 2016 on sleep quality, quality of life, and physical performance among elderly Chinese women with knee
Received in revised form osteoarthritis (OA).
18 March 2016
Methods: A 24-week randomized, controlled trial of 46 elderly women with knee OA. Participants were
Accepted 14 December 2016
Available online 6 February 2017
randomly assigned to either a Tai Ji Quan group (n ¼ 23) or a control group (n ¼ 23). Participants in the
Tai Ji Quan group completed training sessions three times per week, while those in the control group had
bi-weekly educational classes. The primary outcome was total score of the Pittsburgh Sleep Quality of
Keywords:
Tai Chi
Index (PSQI). Secondary outcomes were: seven subscales of the PSQI; sleep latency; total sleep time;
Rheumatic diseases sleep efficiency; physical component summary (PCS) and mental component summary (MCS) of the 36-
Sleep quality item Short Form Health Survey (SF-36); Berg Balance Scale (BBS); and Timed Up and Go (TUG).
Quality of life Results: Compared with the control group, participants in the Tai Ji Quan group had significantly
Aging improved primary outcome (global PSQI score, p ¼ 0.006) and secondary outcomes, including three PSQI
sub-scores (sleep latency, p ¼ 0.031; sleep duration, p ¼ 0.043; daytime dysfunction, p ¼ 0.007), total
sleep time (p ¼ 0.033), and SF-36 PCS (p ¼ 0.006). The Tai Ji Quan group also had significant improve-
ments compared with baseline in three PSQI sub-scores (sleep latency, p ¼ 0.031; habitual sleep effi-
ciency, p ¼ 0.049; sleep disturbance, p ¼ 0.016), sleep latency (p ¼ 0.003), BBS (p ¼ 0.001), and TUG
(p ¼ 0.006).
Conclusion: Tai Ji Quan training is an effective treatment approach to improve sleep quality and quality of
life in elderly Chinese women with knee OA.
Trial registration: Chinese Clinical Trial Registry (June 16, 2013): ChiCTR-TRC-13003264.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction quality of life and mental health [6]. Therefore, sleep quality is an
important outcome relevant to the treatment efficacy of OA [7].
Knee osteoarthritis (OA) is the most common joint disorder and Non-pharmacological interventions, such as exercise, play an
a leading cause of chronic pain, functional limitation and physical important role in the treatment of knee OA [1], and can have
disability; it affects women more frequently than men [1]. The beneficial effects on pain, physical function, sleep quality, and
prevalence of symptomatic knee OA is higher in China than in quality of life [8e10]. Tai Ji Quan is a traditional Chinese exercise
Western countries [2], especially in women (40% higher in Beijing that has the potential to improve muscular strength, joint stability
than in Framingham, USA) [3]. Sleep complaints are more common and flexibility, and reduce pain, depression and physical disability
in people with OA because of chronic pain and poor physical [11,12]. It has been shown that Tai Ji Quan is effective in improving
function [4,5]. Sleep disturbances occur in more than two-thirds of the sleep quality among different populations, including college
OA patients, leading to augmentation of pain severity, and poor students [13], middle-aged adults [14], older adults [15e17], and
cancer survivors [18]. Subjects participating in Tai Ji Quan have
improved self-reported sleep quality, increased sleep duration, and
* Corresponding author. Fax: þ86 21 51253242. reduced sleep-onset latency and sleep disturbances [15,17]. More-
E-mail address: yuliu@sus.edu.cn (Y. Liu). over, there is evidence that this kind of range of motion and low-

http://dx.doi.org/10.1016/j.sleep.2016.12.024
1389-9457/© 2017 Elsevier B.V. All rights reserved.
 J. Lü et al. / Sleep Medicine 33 (2017) 70e75 71

impact physical activity is associated with improvements in regarding health promotion. Health education participants attended
arthritic symptoms and physical function in patients with OA [19]. 78% of scheduled sessions (SD 16%; median [IQR] 77% [67e92%]).
However, most of these prior studies have dealt mainly with pain
and physical function, with relatively little attention being paid to 2.3. Outcome measures
other outcome measures, such as sleep quality and quality of life,
which reflect the real impact of knee OA on patients. All outcome measures were assessed at baseline (before inter-
The primary purpose of this study was therefore to determine vention) and the end of the study (after intervention), and included
the effect of a 24-week Tai Ji Quan training program on self- demographic and clinical profiles, anthropometric measures, study
reported sleep quality among elderly Chinese women with knee primary outcome measures and secondary outcome measures. All
OA. A secondary purpose was to examine the effects of Tai Ji Quan assessments were completed in a research laboratory at Shanghai
on quality of life and physical performance. It was hypothesized University of Sport, and study assessors who conducted outcome
that Tai Ji Quan training would lead to improvement in sleep assessments were blinded to the participants' group allocation.
quality, quality of life, and physical performance outcome
measures. 2.3.1. Primary outcome measures
The Pittsburgh Sleep Quality Index (PSQI, Chinese version [23])
was used as a outcome measure for assessing sleep quality and
2. Methods
disturbances over a one-month time interval, with higher scores
indicating poorer sleep quality [24]. The PSQI is a 19-item self-report
2.1. Study design and participants
scale with seven components: subjective sleep quality, sleep latency,
sleep duration, habitual sleep efficiency, sleep disturbance, use of
The study was designed as a randomized, controlled interven-
sleeping medications, and daytime dysfunction. The global PSQI
tion trial with two groups. Participants were randomly assigned
score is the sum of the seven components, ranging from 0 to 21.
into either a Tai Ji Quan training or health education control group.
The global PSQI score was the primary outcome and the subscale
The study protocol has been described in detail elsewhere [20] and
scores were secondary outcomes in this study. This scale has a high
was approved by the ethics committee of Shanghai University of
degree of internal consistency (Cronbach's alpha ¼ 0.83) with good
Sport. Written informed consent was obtained from all
sensitive (89.6%) and specificity (86.5%) [24].
participants.
Chinese women, who have a higher risk of both OA and age-
2.3.2. Secondary outcome measure
related sleep problems than men [1,21], were recruited from two
Secondary outcome measures included: (1) sleep latency (mi-
community centers in Shanghai between January 2013 and March
nutes), total sleep time (hours) and sleep efficiency (%), and seven
2013. Detailed descriptions of inclusion criteria, recruitment and
subscales of PSQI (range score: 0e3) [24], (2) physical component
adherence have already been presented by Zhu et al. [20]. Briefly:
summary (PCS) and mental component summary (MCS) of the 36-
individuals were eligible if they had a clinical diagnosis of knee OA,
item Short Form Health Survey (SF-36) used as measures of quality
an age of 60e70 years, stable medication use, and willingness to
of life, with a range of 0e100 [25], (3) Berg Balance Scale (BBS)
engage in the study and to be assigned to any of the two in-
assessing balance, which consisted of 14 items, including simple
terventions. Forty-six participants who qualified according to the
mobility tasks (eg, transfers, standing unsupported, sit-to-stand)
eligibility criteria were randomized to the Tai Ji Quan training group
and more difficult tasks (eg, tandem standing, turning 360 ,
(Tai Ji Quan, n ¼ 23) and the health education control group (con-
single-leg stance) [26], (4) Timed Up and Go test (TUG) assessing
trol, n ¼ 23) [20]. Of these, six participants dropped out due to time
mobility in which the participant was asked to stand up from a
conflict and health-related issues (two in the intervention group
chair, walk 3 m, turn around, return to the chair and sit down [27].
and four in the control group).
2.4. Statistical analysis
2.2. Intervention
All analyses were conducted on an intent-to-treat basis so that
Individuals in the Tai Ji Quan group participated in a 60- all participants were included and analyzed regardless of adher-
min session three times a week for 24 weeks. Considering the ence or dropout status. The incomplete data resulting from drop-
symptoms of knee OA and avoiding strenuous activities on joints, outs were handled through the method of “last observation carried
the Tai Ji Quan program was designed to be more suitable for pa- forward” [28]. Baseline characteristics of the study participants
tients with knee OA, and followed an easy-to-difficult progression, were analyzed using analysis of variance for continuous variables
focusing on reducing sustained unilateral weight bearing, dynamic and Chi-squared test or ManneWhitney U test for categorical var-
rotational weight shifting at the knee joints and excessive knee iables. Differences in each group before and after intervention were
flexion. The training program included eight Tai Ji Quan forms analyzed by paired t-test for normally distributed data and Wil-
adapted primarily from the 24-form practice routine [22]: (1) coxon signed-rank test for non-parametric data. Mean changes
“withdraw and push”; (2) “fan through the back”; (3) “wave hands from baseline were compared between groups using repeated
like clouds”; (4) “life hand”; (5) “brush knee and twist steps”; (6) measure analysis of variance. For the variables that were signifi-
“step back to repulse monkey”; (7) “fair lady works at shuttles” and cantly different between groups at baseline, mean changes be-
(8) “golden pheasant stands with one leg (right and left)”. Each tween groups were analyzed using analysis of covariance
exercise session included a 5-min warm-up, 50 min of Tai Ji Quan, (ANCOVA) adjusting for baseline levels of the dependent variables.
and 5-minute cool down, all of which took place in the morning (at Relationships between changes in sleep quality, knee pain, quality
approximately 08:00e09:00). For more details of the Tai Ji Quan of life and physical performance were evaluated using Pearson
intervention, please refer to Zhu et al. [20]. Mean attendance at Tai correlation analysis. An alpha level of 0.05 was considered as sta-
Ji Quan sessions (total of 72) was 87% over 24 weeks (SD 5%; median tistical significance. Data analyses were performed using SPSS
[IQR] 88% [83e91%]). software (version 20.0, IBM Corporation, Armonk, NY, USA).
During the 24-week study period, participants in the control This randomized, controlled trial was originally powered to
group received 60-min bi-weekly wellness education classes examine effects on gait kinematics [20], which determined the
72 J. Lü et al. / Sleep Medicine 33 (2017) 70e75

sample size of 46 participants with 20% attrition rate. A meaningful sleep efficiency, t ¼ 2.09, p ¼ 0.049; sleep disturbance, t ¼ 2.65,
change of PSQI has not been identified. However, this trial had 80% p ¼ 0.016) and sleep latency (t ¼ 3.23, p ¼ 0.003) with 32%
power to detect a moderate effect size of 0.40 (range 0.30e0.80) for improvement, while there was no significant improvement in the
patient-reported outcomes [29], which may be appropriate for control group (Table 2 and Fig. 1).
PSQI, using a two-tailed alpha of 0.05.
3.1.2. Quality of life and physical performance
3. Results There was a significant group difference between the Tai Ji Quan
and control groups for PCS score at 24 weeks (F ¼ 8.42, p ¼ 0.006).
3.1. Baseline characteristics of participants In follow-up pre-to-post-intervention comparisons, participants in
the Tai Ji Quan group demonstrated significant improvement in PCS
Table 1 shows the baseline characteristics of participants, (t ¼ 2.36, p ¼ 0.029), while no significant improvement was shown
including demographic, anthropometric, and clinical features, and in the control group (Table 3).
the outcome measures in the Tai Ji Quan and control groups. There Participants in the Tai Ji Quan group also showed significant pre-
were no significant differences between the two groups (p < 0.05). to-post-intervention improvements in BBS (t ¼ 3.69, p ¼ 0.001) and
TUG (t ¼ 3.05, p ¼ 0.006). Compared with those in the control
3.1.1. Sleep quality group, there was no significant difference between the two groups
At 24 weeks, there was statistically significant difference be- (Table 3).
tween the Tai Ji Quan and control groups for global PSQI score
(F ¼ 8.43, p ¼ 0.006). Participants in the Tai Ji Quan group also 3.1.3. Correlations between changes in sleep quality, pain, quality of
exhibited a significant pre-to-post-intervention improvement in life and physical performance
global PSQI score (t ¼ 3.87, p ¼ 0.001) with a moderate The reduction in PSQI score was significantly associated with
improvement (25%), while no significant improvement was greater reduction in knee pain measured by the Western Ontario
observed in control group. Compared with the control group, par- and McMaster Universities Osteoarthritis Index (WOMAC) score
ticipants in the Tai Ji Quan group showed significant improvements (r ¼ 0.336, p ¼ 0.045; see Zhu et al. [20] for a detailed description of
in three sub-scores of PSQI (sleep latency, F ¼ 4.99, p ¼ 0.031; sleep intervention-related changes) and there was also a significant
duration, F ¼ 4.37, p ¼ 0.043; daytime dysfunction, F ¼ 8.13, negative correlation between changes in sleep latency and SF-36
p ¼ 0.007) and the total sleep time (F ¼ 4.90, p ¼ 0.033). In follow- PCS (r ¼ 0.340, p ¼ 0.037), indicating that improvement in sleep
up pre-to-post-intervention comparisons at 24 weeks, participants quality may be associated with pain relief and better quality of life.
in the Tai Ji Quan group showed significant improvements in three Improvements in sleep quality were not associated with changes in
sub-scores of PSQI (sleep latency, t ¼ 2.32, p ¼ 0.031; habitual physical performance measured by BBS and TUG.

Table 1
Baseline characteristics of the study participants.

Measure Tai Ji Quan (n ¼ 23) Control (n ¼ 23) p

Age, years, Mean (SD) 64.61 (3.40) 64.53 (3.43) 0.758

Body Mass Index, kg/m2, Mean (SD) 25.23 (3.46) 25.05 (3.42) 0.872
Education, years, Mean (SD) 9.86 (1.39) 10.11 (3.35) 0.766
Health statusa (1e5), Mean (SD) 3.48 (0.75) 3.89 (0.81) 0.098
Duration of knee pain, months, Mean (SD) 21.91 (13.62) 22.30 (11.77) 0.937
WOMAC score, Mean (SD)
Knee pain 8.46 (4.90) 9.65 (5.64) 0.457
Knee stiffness 3.57 (2.69) 4.76 (3.00) 0.473
Functional limitation 24.70 (12.95) 25.37 (14.64) 0.267
Radiographic severity of knee, n (%)
K/L Grade 1 7 (30.44) 6 (26.09)
K/L Grade 2 12 (52.17) 14 (60.87)
K/L Grade 3 4 (17.39) 3 (13.04) 0.830c
Comorbidities, n (%)
Neck pain 2 (8.69) 3 (13.04)
Low back pain 4 (17.39) 3 (13.04)
Heart disease 1 (4.34) 0
Hypertension 7 (30.43) 8 (34.78) 0.703c
Sleep quality, Mean (SD)
PSQI (0e21) 6.00 (3.02) 8.47 (4.78) 0.063
Poor sleep quality,b n (%) 9 (39.13) 13 (56.52) 0.376c
Sleep latency, minutes, Median (Range) 30.00 (5, 210) 30.00 (5, 60) 0.510y
Total sleep time, hours 7.00 (1.18) 6.21 (1.46) 0.070
Sleep efficiency (%) 87.27 (14.08) 81.94 (19.56) 0.325
SF-36, Mean (SD)
PCS (0e100) 49.05 (8.25) 44.23 (6.81) 0.056
MCS (0e100) 51.54 (9.93) 50.45 (9.35) 0.727
Physical performance, Mean (SD)
BBS (0e56) 53.05 (3.04) 54.58 (1.74) 0.062
TUG, seconds 8.82 (1.20) 8.62 (1.32) 0.614

SD, Standard Deviation; BMI, Body Mass Index; WOMAC, the Western Ontario and McMaster Universities Osteoarthritis Index; K/L, KellgreneLawrence scale; PSQI, the
Pittsburgh Sleep Quality Index; SF-36, the 36-item Short Form Health Survey; PCS, Physical Component Summary; MCS, Mental Component Summary; BBS, Berg Balance
Scale; TUG, Timed Up and Go test.
c
Chi-squared test.
y
ManneWhitney U test.
a
Measured on a 5-point Likert scale in which 1 indicated poor and 5 indicated excellent.
b
Defined as Global PSQI >5.
 J. Lü et al. / Sleep Medicine 33 (2017) 70e75 73

Table 2
Changes from baseline and differences between groups in sleep quality.

Variable Mean difference from baseline (95% CI); pa Group difference after intervention

Tai Ji Quan Control F pb

Global PSQI (0e21) 1.48 (2.27, 0.68); 0.001 0.47 (2.95, 2.01); 0.693 8.432 0.006
PSQI subscales (0e3)
Subjective sleep quality 0.10 (0.22, 0.41); 0.540 0.21 (0.65, 0.23); 0.331 3.901 0.056
Sleep latency 0.33 (0.63, 0.03); 0.031 0.16 (0.79, 0.47); 0.604 4.989 0.031
Sleep duration 0.14 (0.44, 0.16); 0.329 0.11 (0.35, 0.56); 0.630 4.367 0.043
Habitual sleep efficiency 0.33 (0.67, 0.00); 0.049 0.42 (0.21, 1.05); 0.176 2.128 0.153
Sleep disturbance 0.33 (0.60, 0.07); 0.016 0.21 (0.65, 0.23); 0.331 0.012 0.914
Sleep medicine use 0.05 (0.15, 0.05); 0.329 0.00 (0.43, 0.43); 1.000 1.673 0.204
Daytime dysfunction 0.38 (0.80, 0.04); 0.072 0.42 (1.03, 0.19); 0.163 8.131 0.007z
Sleep latency (minutes) 8.57 (e13.92, 3.23); 0.003y 7.11 (24.30, 10.09); 0.450y 3.747 0.060
Total sleep time (hours) 0.19 (0.20, 0.58); 0.321 0.03 (0.39, 0.44); 0.895 4.897 0.033
Sleep efficiency (%) 5.55 (0.01, 11.11); 0.050 3.13 (8.95, 2.70); 0.274 4.044 0.051

The bold values indicate a statistically significant difference.

PSQI, the Pittsburgh Sleep Quality Index; CI, Confidence Interval.
y
Wilcoxon signed-rank test.
z
Analysis of covariance (ANCOVA) adjusting for baseline level of the variable.
a
Difference between before and after intervention by paired t-test.
b
Between-group difference (Tai Ji Quan and control) and time (before and after intervention) by repeated measure analysis of variance.

Fig. 1. Percent change in global Pittsburgh Sleep Quality Index and other outcomes of sleep quality.

Table 3
Changes from baseline and between-group differences in quality of life and physical performance.

Variable Mean difference from baseline (95% CI); pa Group difference after intervention

Tai Ji Quan Control F pb

SF-36 PCS (0e100) 3.94 (0.45, 7.42); 0.029 2.73 (2.09, 7.54); 0.247 8.415 0.006
SF-36 MCS (0e100) 4.01 (0.17, 8.18); 0.059 0.91 (4.39, 6.21); 0.721 1.851 0.182
BBS (0e56) 1.90 (0.83, 2.98); 0.001 0.00 (1.02, 1.02); 1.000 1.357 0.251
TUG (seconds) 0.73 (1.23, 0.23); 0.006 0.27 (0.77, 0.23); 0.994 0.008 0.931

The bold values indicate a statistically significant difference.

SF-36, the 36-item Short Form Health Survey; PCS, Physical Component Summary; MCS, Mental Component Summary; BBS, Berg Balance Scale; TUG, the Timed Up and Go
test; CI, Confidence Interval.
a
Difference between before and after intervention by paired t-test.
b
Between-group difference (Tai Ji Quan and control) and time (before and after intervention) by repeated measure analysis of variance.
74 J. Lü et al. / Sleep Medicine 33 (2017) 70e75

4. Discussion contributing factor behind the positive effect of Tai Ji Quan on sleep
quality in patients with knee OA [37]. The present study is in
The primary purpose of this study was to clarify the effects of Tai keeping with these findings, since, as reported elsewhere, pain
Ji Quan on self-reported sleep quality in elderly Chinese women intensity was reduced by 39% in the Tai Ji Quan group [20], and a
with knee OA. The major findings were that elderly women with significant correlation between reductions in pain and PSQI total
knee OA benefited from 24-week Tai Ji Quan training, while no score (r ¼ 0.336) was observed in the present study. Second, sig-
improvement was observed in the health education control group. nificant improvements in quality of life and physical performance
Compared with the control group, participants in the Tai Ji Quan from Tai Ji Quan training were found in the present study, and these
group also showed a concomitant improvement in quality of life factors have previously been associated with sleep quality in OA
measures. The findings of this randomized controlled trial suggest patients [4,37]. There was also a significant negative relationship
that 24-week Tai Ji Quan training is an effective treatment for elderly between changes in sleep latency and SF-36 PCS (r ¼ 0.340).
women with knee OA that improves sleep quality and quality of life. Finally, the practice of Tai Ji Quan is a mind-body exercise
In the present study, global PSQI score showed a significant comprising physical movements, but also with a psychological
decrease by, on average, 1.42 points in elderly women with knee OA focus and emphasis on relaxation, mindfulness of movement and
participating in Tai Ji Quan, representing a moderate degree of breathing that may influence depression, stress, anxiety, emotion,
improvement (25%). Significant group differences were also and self-esteem, all of which may be considered as potential me-
observed in total sleep time and three subscale scores of PSQI (sleep diators of improved sleep quality.
latency, sleep duration, and daytime dysfunction). Although no Major strengths of this study included the randomized clinical
between-group difference was observed, there were significant trial design, the comprehensive evaluation of sleep quality, and the
within-group changes (improvements) in sleep latency, sleep effi- use of a Tai Ji Quan protocol designed specifically for knee OA.
ciency and other two subscale scores of PSQI (habitual sleep effi- However, there were some potential limitations of the study. First,
ciency and sleep disturbance). These results are generally the results of the intervention effects might have been limited by
congruent with the findings of other studies, although the samples the relatively small sample size and the fact that only female sub-
used in the earlier studies were not limited to persons with knee jects were included. This may have restricted the generalizability of
OA. Two studies by Li et al. [15] and Irwin et al. [17], which the results concerning Tai Ji Quan benefits. Second, outcomes of
examined the effects of Tai Chi on elderly adults, found significant sleep quality and quality of life in this study were measured by
improvements among participants engaging in Tai Chi training. subjective questionnaires (PSQI and SF-36) and it was impossible to
Nguyen et al. [16] examined the effect of Tai Chi on sleep quality in blind the participants to treatment. Therefore, the subjective
healthy elder Vietnamese. These studies found significantly measures might have introduced a bias, leading to potential over-
reduced PSQI global score (decreased by 1.8e2.06 points) and the estimation of the intervention effects. Third, due to the absence of
subscale scores, such as subjective sleep quality, sleep duration, follow-up beyond the 24-week intervention period, it was unclear
sleep efficiency and sleep disturbance [15-17]. Moreover, Larkey whether Tai Ji Quan training has lasting health effects on in-
et al. [30] demonstrated that Tai Chi training also had positive ef- dividuals with knee OA.
fects on sleep quality and fatigue among cancer survivors. In conclusion, this study suggests that Tai Ji Quan training is an
Although the sleep benefits from Tai Ji Quan have been well effective and safe treatment for knee OA that improves sleep quality
demonstrated [15e17,30], it is believed that this current study is the and quality of life. Given the promising results, future studies are
first to evaluate improvements in sleep quality from a Tai Ji Quan needed to demonstrate the generalizability of Tai Ji Quan benefits in
intervention specifically in knee OA patients. The increased joint clinical populations with knee OA, and using a longer follow-up
pain and disability associated with knee OA could have direct time to examine the sustainability of these effects.
impact on other aspects of subjects' lives, including sleep quality,
quality of life, emotional status, stress and anxiety [31,32]. Acknowledgement
Compared with subjects with other chronic diseases or healthy
controls, subjects with OA experience more frequent and more This study was supported by the National Natural Science
significant sleep problems [4,6,33]. Poor sleep quality is not only a Foundation of China (81572213) and the Science and Technology
consequence of chronic pain in OA, it is also likely to play an inte- Commission of Shanghai (14DZ1103500).
gral role in augmenting clinical OA pain [34]. Therefore, aggressive
management of sleep quality could be an important approach to
Conflict of interest
ensuring effectiveness of the treatment and maintaining a higher
level of quality of life in individuals with OA [7].
The authors declare that there is no potential conflict of interest
In addition to demonstrating improvement in sleep quality
with this manuscript.
measures, participants in the Tai Ji Quan group also showed sig-
The ICMJE Uniform Disclosure Form for Potential Conflicts of
nificant improvements in quality of life (SF-36 PCS) and physical
Interest associated with this article can be viewed by clicking on the
performance (BBS and TUG). Although statistically significant
following link: http://dx.doi.org/10.1016/j.sleep.2016.12.024.
changes were observed in these measures after the 24-week
intervention, the magnitude of the changes was modest (4e8%)
and much lower than previously reported in other studies [35,36]. References
This lower degree of improvement may be due to differences in [1] Felson DT. Osteoarthritis of the knee. N Engl J Med 2006;354:841e8.
style, duration and intensity of the Tai Ji Quan intervention and the [2] Tang X, Wang S, Zhan S, et al. The prevalence of symptomatic knee osteoar-
characteristics of the participants. thritis in China: results from China health and retirement longitudinal study.
Arthritis Rheumatol 2015;68:648e53.
There are several factors that may explain the positive effects of
[3] Zhang Y, Xu L, Nevitt MC, et al. Comparison of the prevalence of knee oste-
Tai Ji Quan training on sleep quality in elderly Chinese women, oarthritis between the elderly Chinese population in Beijing and whites in the
although the mechanism by which Tai Ji Quan induces such United States: the Beijing Osteoarthritis Study. Arthritis Rheum 2001;44:
improvement of sleep quality is not fully understood. First, the ef- 2065e71.
[4] Chen CJ, McHugh G, Campbell M, et al. Subjective and objective sleep quality
ficacy of Tai Ji Quan on relieving pain symptoms among patients in individuals with osteoarthritis in Taiwan. Musculoskelet Care 2015;13(3):
with OA has been well demonstrated [19], and this may be a major 148e59.
 J. Lü et al. / Sleep Medicine 33 (2017) 70e75 75

[5] Hawker GA, French MR, Waugh EJ, et al. The multidimensionality of sleep [21] Foley D, Ancoli-Israel S, Britz P, et al. Sleep disturbances and chronic disease in
quality and its relationship to fatigue in older adults with painful osteoar- older adults: results of the 2003 National Sleep Foundation Sleep in America
thritis. Osteoarthr Cartil 2010;18:1365e71. Survey. J Psychosom Res 2004;56:497e502.
[6] Pickering ME, Chapurlat R, Kocher L, et al. Sleep disturbances and osteoar- [22] Guo Y, Qiu P, Liu T. Tai Ji Quan: an overview of its history, health benefits, and
thritis. Pain Pract 2016;16:237e44. cultural value. J Sport Health Sci 2014;3:3e8.
[7] Turk DC, Cohen MJ. Sleep as a marker in the effective management of chronic [23] Liu X, Tang M, Hu L, et al. Reliability and validity of the Pittsburgh sleep
osteoarthritis pain with opioid analgesics. Semin Arthritis Rheum 2010;39: quality index. Chin J Psychiatry 1996;29:103e7.
477e90. [24] Buysse DJ, Reynolds CF, Monk TH, et al. The Pittsburgh Sleep Quality Index: a
[8] Rejeski WJ, Mihalko SL. Physical activity and quality of life in older adults. new instrument for psychiatric practice and research. Psychiatry Res 1989;28:
J Gerontol A Biol Sci Med Sci 2001;56:23e35. 193e213.
[9] O'Reilly SC, Muir KR, Doherty M. Effectiveness of home exercise on pain and [25] Ware JE, Kosinski M, Dewey JE, et al. SF-36 health survey: manual and
disability from osteoarthritis of the knee: a randomised controlled trial. Ann interpretation guide. Lincoln, RI: Quality Metric Inc; 2000.
Rheum Dis 1999;58:15e9. [26] Berg K, Wood-Dauphine S, Williams J, et al. Measuring balance in the elderly:
[10] Taibi DM, Vitiello MV. A pilot study of gentle yoga for sleep disturbance in preliminary development of an instrument. Physiother Can 1989;41:304e11.
women with osteoarthritis. Sleep Med 2011;12:512e7. [27] Mathias S, Nayak U, Isaacs B. Balance in elderly patients: the “get-up and go”
[11] Wang C, Collet J, Lau J. The effect of Tai Chi on health outcomes in patients test. Arch Phys Med Rehabil 1986;67:387e9.
with chronic conditions. Arch Intern Med 2004;13:12e3. [28] Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guide-
[12] Li G, Yuan H, Zhang W. Effects of Tai Chi on health related quality of life in lines for reporting parallel group randomised trials. BMC Med 2010;8:1.
patients with chronic conditions: a systematic review of randomized [29] Cella D, Eton DT, Lai JS, et al. Combining anchor and distribution-based
controlled trials. Complement Ther Med 2014;22:743e55. methods to derive minimal clinically important differences on the Func-
[13] Caldwell K, Harrison M, Adams M, et al. Effect of Pilates and Taiji Quan tional Assessment of Cancer Therapy (FACT) anemia and fatigue scales. J Pain
training on self-efficacy, sleep quality, mood, and physical performance of Symptom Manage 2002;24:547e61.
college students. J Bodyw Mov Ther 2009;13:155e63. [30] Larkey LK, Roe DJ, Weihs KL, et al. Randomized controlled trial of Qigong/Tai
[14] Lee LY, Tam KW, Lee ML, et al. Sleep quality of middle-aged Tai Chi practi- Chi Easy on cancer-related fatigue in breast cancer survivors. Ann Behav Med
tioners. Jpn J Nurs Sci 2015;12:27e34. 2015;49:165e76.
[15] Li F, Fisher KJ, Harmer P, et al. Tai chi and self-rated quality of sleep and [31] Ferrell BA. Pain management in elderly people. J Am Geriatr Soc 1991;39:
daytime sleepiness in older adults: a randomized controlled trial. J Am Geriatr 64e73.
Soc 2004;52:892e900. [32] Alves JC, Bassitt DP. Quality of life and functional capacity of elderly women
[16] Nguyen MH, Kruse A. A randomized controlled trial of Tai chi for balance, with knee osteoarthritis. Einstein (Sao Paulo, Brazil) 2013;11:209e15.
sleep quality and cognitive performance in elderly Vietnamese. Clin Interv [33] Wilcox S, Brenes GA, Levine D, et al. Factors related to sleep disturbance in
Aging 2012;7:185. older adults experiencing knee pain or knee pain with radiographic evidence
[17] Irwin MR, Olmstead R, Motivala SJ. Improving sleep quality in older adults of knee osteoarthritis. J Am Geriatr Soc 2000;48:1241e51.
with moderate sleep complaints: a randomized controlled trial of Tai Chi Chih. [34] Smith MT, Quartana PJ, Okonkwo RM, et al. Mechanisms by which sleep
Sleep 2008;31:1001e8. disturbance contributes to osteoarthritis pain: a conceptual model. Curr Pain
[18] Fong SS, Ng SS, Lee HW, et al. The effects of a 6-month Tai Chi Qigong training Headache Rep 2009;13:447e54.
program on temporomandibular, cervical, and shoulder joint mobility and [35] Wang C, Schmid CH, Hibberd PL, et al. Tai Chi is effective in treating knee
sleep problems in nasopharyngeal cancer survivors. Integr Cancer Ther osteoarthritis: a randomized controlled trial. Arthritis Rheum 2009;61:1545e53.
2015;14:16e25. [36] Lee H-J, Park H-J, Chae Y, et al. Tai Chi Qigong for the quality of life of patients
[19] Yan JH, Gu WJ, Sun J, et al. Efficacy of Tai Chi on pain, stiffness and function in with knee osteoarthritis: a pilot, randomized, waiting list controlled trial. Clin
patients with osteoarthritis: a meta-analysis. PloS One 2013;8:e61672. Rehabil 2009;23:504e11.
[20] Zhu Q, Huang L, Wu X, et al. Effects of Tai Ji Quan training on gait kinematics in [37] Taylor-Gjevre RM, Gjevre JA, Nair B, et al. Components of sleep quality and
older Chinese women with knee osteoarthritis: a randomized controlled trial. sleep fragmentation in rheumatoid arthritis and osteoarthritis. Musculoskelet
J Sport Health Sci 2016;5(3):297e303. Care 2011;9:152e9.