Tai Chi for Chronic Pain Conditions: A Systematic Review and Meta-analysis of Randomized Controlled Trials

Several studies reported that Tai Chi showed potential effects for chronic pain, but its role remains controversial. This review assessed the evidence regarding the effects of Tai Chi for chronic pain conditions. 18 randomized controlled trials were included in our review. The aggregated results have indicated that Tai Chi showed positive evidence on immediate relief of chronic pain from osteoarthritis (standardized mean difference [SMD], −0.54; 95% confidence intervals [CI], −0.77 to −0.30; P < 0.05). The valid duration of Tai Chi practice for osteoarthritis may be more than 5 weeks. And there were some beneficial evidences regarding the effects of Tai Chi on immediate relief of chronic pain from low back pain (SMD, −0.81; 95% CI, −1.11 to −0.52; P < 0.05) and osteoporosis (SMD, −0.83; 95% CI, −1.37 to −0.28; P = 0.003). Therefore, clinicians may consider Tai Chi as a viable complementary and alternative medicine for chronic pain conditions.


Statistical analysis.
For the continuous data, the changes from baseline were used in the meta-analysis. The pain outcomes were presented as the standardized mean difference (SMD) and 95% confidence intervals (CI) because the scales were not consistent across eligible studies. The meta-analyses were conducted based on at least two trials using Review Manager Version 5.0. For the expected heterogeneity, the continuous data were pooled using a more conservative random-effects model. The heterogeneity was assessed using the Cochran Q statistic (P < 0.10, was considered to be statistically significant) and was quantified using the I 2 index (where I 2 > 30% indicated moderate heterogeneity; I 2 > 50% substantial heterogeneity; and I 2 > 75% considerable heterogeneity). P < 0.05 was considered to be statistically significant.
The subgroup analysis was conducted based on different diseases. If the studies had two or more control groups, the following order was used to select the control group: placebo; no treatment; waiting list control; attention control; education; and other active interventions. To identify the effects of Tai Chi for chronic pain of OA, the subgroup analyses were performed according to different control interventions and intervention durations of Tai Chi exercise. The publication bias was assessed using funnel plots.

Results
Study selection. A total of 706 records were identified from 7 English and Chinese databases. After removing the duplicates, 217 potentially relevant abstracts were initially screened, and 188 were excluded for failing to meet the inclusion criteria. We retrieved and reviewed 29 full-text articles. A total of 18 RCTs were eligible for this review, including 15 English articles 9-12,20-30 and 3 Chinese articles [31][32][33] , as indicated by the flowchart in Fig. 1. During the screening full-texts, the studies were excluded for the following reasons: they were not randomized (n = 5) 34-38 ; they were duplicate publications (n = 1) 39 ; they were suspected of being counterfeit (n = 1) 40 ; and they failed to present the available data (n = 4) 41-44 . Characteristics of included studies. A total of 1260 individuals with a mean age of 61.43 ± 10.99 years were included in eligible RCTs, which were conducted in Australia, China, Korea, and the USA between 2000 and 2015. The duration of the included studies was between 6 weeks and 28 weeks. The mean ± standardized difference of the therapeutic session and time were 50.44 ± 49.62 (range 10-168 minutes) and 54.72 ± 14.80 minutes (range 20-90 minutes). The follow-up time ranged from 6 weeks to 36 weeks. Of all the studies, 8 RCTs assessed the effectiveness of Tai Chi for OA 10,11,20,21,23,24,27,28 , 3 for LBP 9,31,33 , 2 for osteoporosis 22,32 , 2 for fibromyalgia 12,26 , and 3 for other diseases 10,25,30 . Seven used the Yang style 11,12,21,22,26,29 , 3 practiced the Sun style 10,23,28 and 1 used the Wu style 20 . The control groups were conducted in attention control, waiting list control, education, routine treatment control, and other active intervention controls including physical therapy and hydrotherapy. The main characteristics of all included RCTs are shown in Table 1. Table 2, the majority (94%) of the included trials exceeded the predetermined cutoff score of 6, ranging from 5 to 8 points for OA 10,11,20,21,23,24,27,28 , LBP 9,31,33 , and fibromyalgia 12,26 , 6 to 7 points for osteoporosis 22,32 , 6 points for herpes zoster 25 , 8 points for RA 29 , and 6 points for stroke 30 . The most common flaws were that the subjects and therapists in all of the trials were unblinded to the treatments, and that 7 RCTs did not perform assessors-blinding 20,25,28,[30][31][32][33] . Additionally, allocation concealments were unclear because the detailed allocation procedure was not reported in 8 trials 21,[24][25][26][30][31][32][33] . The intention-to-treat analysis was rated positive in 12 studies [9][10][11][12]20,23,25,27,29,[31][32][33]  Three trials reported the follow− up effects of Tai Chi for OA chronic pain conditions 11,21,23 . Two RCTs assessed the short term effects of Tai Chi after a 6-week follow-up 21 and a 12-week follow-up 11 . Although one of them reported that Tai Chi was effective after a 12-week follow-up, the aggregated results indicated that Tai Chi did not show better short term effects (SMD, −0.26; 95% CI, −1.04 to 0.51; P = 0.51; Fig Tai Chi for fibromyalgia. Two RCTs tested the effects of Tai Chi for fibromyalgia 12,26 . Although one trial reported better effects of Tai Chi than education and stretching 12 , the aggregated results did not support better effects of Tai Chi in improving fibromyalgia pain (SMD, −0.52; 95% CI, −1.09 to 0.05; P = 0.07; Fig. 2) 12,26 . One trial reported that Tai Chi showed better short term follow-up effects for fibromyalgia pain after a 12-week follow-up (mean changes, 2.4 versus 0.7, P < 0.05) 12 . Tai Chi for other diseases. Three studies tested the effects of Tai Chi for herpes zoster (postherpetic pain) 25 , RA 29 , and stroke 30 . Irwin et al. reported that Tai Chi exercise showed significant improvements in body pain from herpes zoster compared with health education (mean changes, 6.68 versus 3.79, P < 0.05) 25 . The study supported that Tai Chi achieved improvements of pain in patients with chronic stroke compared to general physical therapy (mean changes, 5.55 versus 0.82, P < 0.05) 30 .

Methodological quality. As shown in
One trial reported that Tai Chi significantly improved RA pain compared with attention control (mean changes, 1.00 versus − 1.60, P < 0.05) 29 . It assessed the short term follow-up effects of Tai Chi for chronic RA pain and reported that 90% of the patients experienced improvements compared to the baseline in joint pain after a 12-week follow-up.

Adverse events.
Only 2 studies reported that there were minor adverse events 9,21 . One study found sporadic complaints of minor muscle soreness and foot and knee pain at the commencement of the intervention 21 . The other study reported that three participants found a small initial increase in back pain symptoms that were alleviated by the third or fourth week of treatment, and one participant reported an increase in upper back pain that was alleviated once the upper extremity posture had been corrected 9 .

Discussion
The major purpose of the current review was to evaluate the effects of Tai Chi for chronic pain conditions. The primary finding was that Tai Chi showed improvements in chronic pain for patients with OA, LBP, and osteoporosis. The valid duration of Tai Chi for chronic OA pain might be at least 6 weeks. On the follow-up effects, there was insufficient evidence of the effects of Tai Chi for suffers of chronic pain conditions. This systematic review assessed the effects of Tai Chi on chronic pain in various common diseases including OA, LBP, RA, osteoporosis, and fibromyalgia. Therefore, a subgroup analysis was performed based on different diseases. The results maintained that Tai Chi showed better effects in improving chronic pain caused by OA, LBP, and osteoporosis, however, there was only moderate evidence of the effects of Tai Chi on chronic pain in patients with OA because other aggregated results were based on fewer eligible studies. Furthermore, the subgroup analysis was performed to compare Tai Chi with different control interventions for chronic pain in patients with OA. The aggregated results indicated that Tai Chi was more effective for participants with chronic OA pain than for those in the waiting list control or attention control groups. However, there was insufficient evidence to support or refute the value of Tai Chi compared with other active therapies because Tai Chi was compared with physical therapy and hydrotherapy in only 2 studies each. Additional, further studies should compare Tai Chi with more active therapies, such as aerobic exercise and acupuncture. Few studies have investigated the follow-up effects of Tai Chi for chronic pain conditions. As for chronic and recurrent pain, more attention should be focused on the long term effects of Tai Chi exercises.
Our results indicated that a minimal valid duration of Tai Chi for chronic pain might be 6 weeks, and the longer duration may achieve better gains. A subgroup analysis supported that 6-10 weeks of Tai Chi significantly improved chronic pain in patients with OA and that long term Tai Chi (12-20 weeks) may be more effective, which is consistent with previous findings 14 . Furthermore, 10-28 weeks of Tai Chi also showed greater improvements in patients with chronic pain of RA, LBP, and osteoporosis. Consequently, long term Tai Chi exercise could be more effective for the management of chronic pain conditions.
Our results are similar to the latest systematic review. Peng's systematic review suggested that Tai Chi seemed to be an effective intervention in OA, LBP and fibromyalgia, however, it was only a qualitative review including 10 RCTs published between 2000 and 2011 17 . Any qualitative reviews may be problematic because they are often more subjective than quantitative meta-analyses. Two studies were excluded in Peng's review because one was not a formal published dissertation 20 and the other used Tai Chi Qigong as the intervention 27 . However, "Tai Chi Qigong" only included Tai Chi exercise. Thus, they were eligible studies for our review. Furthermore, detailed subgroup analyses were performed based on different diseases, control interventions, and durations of Tai Chi. Additionally, the follow-up effects of Tai Chi for chronic pain conditions were focused in our review. Therefore, there was more powerful evidence of Tai Chi for chronic pain conditions in our review.

Study limitations.
There are several limitations in our review: (a) A rigorous search strategy was applied in our review, however, some uncertainty still remains due to bias in location and publication 45,46 . (b) Although the predetermined cutoff score of 6 using PEDro scale was exceeded by the majority of studies, there were some flaws in the blinding methods of eligible RCTs. It is difficult to blind the patients and it is impossible to blind the therapists in Tai Chi studies, however, the blinded assessors and concealed allocations should compensate for these flaws. Several trials did not perform these compensated methods. It was suspicious that no participant dropped out during the Tai Chi intervention that lasted for at least 6 months in three Chinese RCTs [31][32][33] . These flaws may       have created potential performance biases and detection biases. Thus, several studies could not be considered to be of high quality. (c) Few eligible RCTs were a major limitation, especially for RA, fibromyalgia, herpes zoster, stroke, osteoporosis, and LBP. Some subgroup analyses were only based on 2 to 3 studies; thus, some conclusions should be interpreted with caution. (d) Our results may be affected by the styles and dosing parameters of Tai Chi such as different styles (Yang-style, Wu-style, etc.) and frequency (number of Tai Chi sessions per week). The eligible studies employed different styles and dosing parameters. (e) Although the pain outcomes were presented as SMD in the meta-analyses, the aggregated results may also be influenced by different outcome measures in eligible studies. Thus, the reliable and valid outcome measure is essential to reduce bias, provide precise measures and perform valid data synthesis. (f) Although fewer adverse events were associated with Tai Chi, definite conclusions are not possible. It can only be assumed that Tai Chi is a treatment option with a low risk of injury.

Conclusion
This systematic review demonstrated positive evidence regarding the effects of Tai Chi on chronic OA pain, and some beneficial evidences of Tai Chi for LBP and osteoporosis. The minimal valid duration of Tai Chi for chronic OA pain may be 6 weeks, and a longer duration of Tai Chi exercise may achieve more gains. However, there was no valid evidence on the follow-up effects of Tai Chi for chronic pain conditions. There was insufficient evidence to support or refute the value of Tai Chi compared with other active therapies for chronic pain conditions. Consequently, future studies should emphasize high-quality RCTs comparing Tai Chi with other active therapies for chronic pain conditions, and a long term follow-up should be conducted.