Tai Chi for Schizophrenia: A Systematic Review.

BACKGROUND: Tai Chi as a form of moderate aerobic exercise originating in China, could promote balance and healing of the mind-body. Furthermore, Tai Chi has been used as an adjunctive treatment for patients with schizophrenia. However, no meta-analysis or systematic review on adjunctive Tai Chi for patients with schizophrenia has yet been reported. AIM: A systematic review and meta-analysis was conducted to examine the efficacy of Tai Chi as an adjunctive treatment for schizophrenia using randomized controlled trial (RCT) data.
METHOD: Two evaluators independently and systematically searched both English- and Chinese-language databases for RCTs of Tai Chi for schizophrenia patients, selected studies, extracted data, conducted quality assessment and data synthesis. Statistical analyses were performed using the Review Manager (version 5.3). The Cochrane Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) was used to assess the strength of the evidence.
RESULTS: In 6 RCTs conducted in mainland China and Hong Kong, there were 483 participants including 215 subjects in the intervention group and 268 subjects in the control group. The trials lasted 16.0 (6.2) weeks. Compared to control group, we found significant differences regarding improvement of negative symptoms assessed by the Positive and Negative Syndrome Scale (PANSS) negative symptom sub-score (2 trials) and Scale for the Assessment of Negative Symptoms (SANS) (3 trials) over the study period in the intervention group (5 trials with 6 treatment arms, n=451, SMD: -0.87 (95%CI: -1.51, -0.24), p=0.007; I2=90%). Furthermore, there is no significant difference regarding improvement of positive symptoms assessed by the PANSS positive symptom sub-score (2 trials) and Scale for the Assessment of Positive Symptoms (SAPS) (2 trials) over the study period (4 trials with 5 treatment arms, n=391, SMD: -0.09 (95%CI: -0.44, 0.26), p=0.60; I2=65%). All included RCTs did not report side effects. Based on the GRADE, the strength of the evidence for primary outcome was 'very low'.
CONCLUSIONS: The data available on the effectiveness of adjunctive Tai Chi in patients with schizophrenia who are receiving antipsychotic is insufficient to arrive at a definitive conclusion about its efficacy. Furthermore, follow-up time in the available studies was relatively short, and all studies did not use blinded assessment of outcome measures. High-quality randomized trials are needed to inform clinical recommendations.

1. Introduction

Apart from psychotic symptoms, schizophrenia patients often experience increased risk of diabetes, hypertension, cardiovascular disease and even mortality due to adverse drug reactions, poor lifestyles or lack of exercise, which will further contribute to worse recovery prognosis and shorter life expectancies[. A large number of recent studies indicated that aerobic exercise can significant improve psychiatric symptoms and increase quality of life[.

Tai Chi as a form of moderate aerobic exercise originating in China, could promote balance and healing of the mind-body[. Furthermore, Tai Chi comprises mental concentration, physical balance, relaxed breathing, and muscle relaxation, which shows great benefits in the rehabilitation of many medical and psychological conditions[, including depression[, rheumatoid arthritis[, Parkinson’s disease[, cancer[,chronic obstructive pulmonary disease [, type 2 diabetes [, and osteoarthritis [.

However, there is lack of evidence from systematic review or meta-analysis supporting the effects of Tai Chi on schizophrenia although it has been commonly used as adjunctive intervention to treat this disorder in Chinese societies[. Thus, we conducted a systematic review and meta-analysis including the recent Tai Chi randomized controlled trials (RCTs) for patients with schizophrenia from both English and Chinese database, the latter being not widely known to the international readership to assess the efficacy of Tai Chi as an intervention treatment.

2. Methods

2.1 Types of studies

Any RCTs reporting the outcomes of the effects of Tai Chi for schizophrenia patients were includedin this review. We excluded case series, non-randomized studies, animal trials, and literature reviews.

2.2 Types of participants

Patients with a clinical diagnosis of schizophrenia, who met any diagnostic criteria (i.e. Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV)[, International Classification of Diseases, 10th edition (ICD-10)[, and Chinese Classification of Mental Disorders, third edition (CCMD-3)[ were included.

2.3 Types of interventions

Interventions were limited to Tai Chi excluding other psychological interventions such as yoga, meditation, and mixed interventions (e.g., Tai Chi with music therapy or Tai Chi with acupuncture), and comparisons with any treatment (e.g. yoga, waitlist, and usual daily activities, et al) were included.

2.4 Types of Outcome measures

The primary outcome psychopathology measure was both positive and negative symptoms assessed by the positive and negative syndrome scale (PANSS)[, Scale for the assessment of negative (SANS)[, or Scale for the Assessment of Positive Symptoms (SAPS)[, respectively. Key secondary outcome were social function, abnormal behaviors, cognitive function, perceived stress, adverse events and discontinuation rate. Furthermore, we recorded clinical outcomes based on intent to treat (ITT) analysis if available.

2.5 Data Searches

PubMed, PsycINFO, Embase, Cochrane Library databases, the Cochrane Controlled Trials Register and Chinese databases (WanFang Database, Chinese Biomedical database and China Journal Net) were searched, from inception of the database until Aug 27 2016, using the following search terms: (Schizophrenia OR Schizophrenias OR Schizophrenic Disorder OR Disorder, Schizophrenic OR Disorders, Schizophrenic OR Schizophrenic Disorder OR Dementia Praecox) AND (Tai-ji OR Tai Chi OR Chi, Tai OR Tai Ji Quan OR Ji Quan, Tai OR Quan, Tai Ji OR Taiji OR Taijiquan OR T’ai Chi OR Tai Chi Chuan) AND (randomized controlled trial OR controlled clinical trial OR randomized OR placebo OR randomly OR trial OR groups).

We also hand-searched reference lists from identified and relevant review articles for additional studies, and contacted authors for unpublished data.

2.6 Data extraction

Two of the authors (WZ and TG) independently conducted the original research, study selection, data extraction, quality assessment, and data synthesis. Any inconsistencies were resolved by discussion. Data were extracted into simple, standardized forms. Furthermore, authors were contacted to obtain missing information or clarification.

2.7 Statistical methods

The statistical analyses were performed using Review Manager (version 5.3) in this meta-analysis according to the recommendations of the Cochrane Collaboration. To minimize the effect of chance, we only meta-analyzed outcomes for which data were available in at least three studies. For continuous data, the effect size (ES) was determined by calculating weighted mean difference (WMD) or standard mean difference (SMD) with 95% confidence interval (CI) according to outcome measurements; for dichotomous data risk ratio (RR)with 95% CI was computed. We assessed the Q-test and I2 to quantify heterogeneity: if p>0.10 or I2<50, it indicates that the study results were homogeneous[ and a fixed effect model was employed to meta-analyze the pooled sample; otherwise a random effects model was employed[. Furthermore, one RCT[ with three study arms was included in the meta-analysis. In order to include each of the 2 adjunctive Tai Chi arms separately, we included the Tai Chi treatment arm twice in the analysis, but assigned half of the patient group to each arm in order to not inflate the total number of patients. Publication bias was assessed using funnel plots and Egger’s test[. All analyses were 2 tailed, with alpha set at 0.05.

2.8 Risk of bias assessment

The Cochrane Risk of Bias including the methods of random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and other biases was used to assess the methodological quality of RCTs [39].

2.9 Clinical evidence recommendation

The quality of evidence and strength of recommendations of outcome measure of Tai Chi for schizophrenia was assessed using the grading of recommendations assessment, development, and evaluation (GRADE) system as ‘very low’, ‘low’, ‘moderate’, or ‘high’[40].

3. Results

3.1 Results of the search

Figure 1 shows that 6 RCTs (n=483)[ from Chinese (4trials)[and English databases (2 trials)[ were included in this meta-analysis after removing duplicate articles (n=17), reviewing the titles or abstracts (n=29) and full texts (n=3) from originally relevant articles (n=55).

Figure 1.

Identification of included studies

3.2 The characteristics of included studies

Table 1 summarizes6 RCTs[conducted in Chinese societies. A total of 483 participants (range: 30-153, median 70) including 215 subjects in the intervention group and 268 subjects in the control group lasted 16.0 (6.2) (range: 12-24, median 12) weeks. The participants with available data were 41.3 (10.4) (range 18-65, median 38.9) years old, 64.0 (20.5)% (range 40%-100%, median 59.5%) was males, and the mean illness duration was 13.8 (12.1) (range=2.6-29.9, median=9.7) years.

Table 1.

Studies and Patients Characteristics

Study	N	Design:-Blinding-Setting	TrialDuration(weeks)	Country	Participants:-Diagnosis-Criteria-Illness duration (year)	Ageb:years(range)	Sex:Male(%)	Interventions:Number of patients	Intervention frequency (Tai Chi)	Outcomes	Treatment adherence (Tai Chi, %)	Dropout rate (Tai Chi, %)
Xie et al 200811	100	-Open label-Inpatients	12	China	-Sz-CCMD-3 -Illness duration: 10.3	35.6 (18-60)	n=60 (60%)	1. APs + TC + routine care; n=502. APs + routine care; n=50	45 min, 3×/wk (24 style)	IPROS;SANS;SAPS	98	2
Gan et al 2007[10]	80	-Open label-Inpatients	12	China	-Sz-CCMD-3-Illness duration: 2.6	27.2 (19-55)	n=47 (59%)	1. APs + TC; n=392. APs + without any kind of exercise; n=41	60min,5×/wk(24 style)	PANSS;NOSIE;SDSS	100	0
Zhou et al 2011[12]	60	-Open label -Inpatients	12	China	-Sz-CCMD-3-Illness duration:2.8	40.2 (21-62)	n=43 (72%)	1. APs + TC; n=302. APs + health education; n=30	60min, 7×/wk (24 styl)	SANS;TRS	100	0
Ho et al 2012[4]	30	-Open label-Othersa	12	China	-SzDSM-IV-Illness duration 27.8	53.0 (18-65)	n=12 (40%)	1. APs + TC; n=152. APs + waitlist; n=15	60min, 2×/wk+30 min, l×/wk (Wu-style)	CMDT;SNS;W H 0- DAS-II	NR	20
Chen et al 2013[13]	60	-Open label-Inpatients	24	China	-Sz-CCMD-3-Illness duration 9.1	37.6 (26-60)	n=60 (100%)	1. APs + TC; n=30 2. APs + + without any kind of exercise; n=30	60min, 7×/wk (Sun-style)	SANS;SAPS;BMT;CPT	100	0
Ho et al 2016[5]	153	-Open label-Othersa	24	China	-Sz-DSM-IV-Illness duration 29.9	54.0 (18-65)	n=80 (53%)	1. APs + TC; n=53. APs + exercise program; n=51	60min, 1 × / wk+45min, 2x/wk (Wu-style)	PANSS;NES;AD- L;WAIS-III;IADL;PSS	NR	2

apatients come from both long-stay care and halfway house services;

bweighted mean; APs = antipsychotics; ADL = Activities of Daily Living; BMT=Backward Masking Test; CMDT = Movement Coordination Tests; CCMD-3 = Chinese Classification of Mental Disorders, third edition; CPT = continuous performance test; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders 4th edition; IPROS = Inpatient Psychiatric Rehabilitation Outcome Scale; IADL = Lawton’s Instrumental Activities of Daily Living; NR = not reported; NOSIE = Nurse’s Observation Scale for Inpatient Evaluation; NES = Neurological Evaluation Scale; PSS = Perceived Stress Scale; PANSS = Positive and Negative Syndrome Scale; SDSS = Social Disability Screening Schedule; R= range; SANS = Scale for the Assessment of Negative Symptoms; SAPS = Scale for the Assessment of Positive Symptoms; Sz = schizophrenia; TC = Tai Chi; TRS=Treatment Rehabilitation Scale; WHODAS-II = World Health Organization Disability Assessment Schedule; WAIS-III = Wechsler Adult Intelligence Scale, Third Edition–Chinese version.

Five RCTs[ implemented two arms and one[ implemented a three-arm design in which a second comparison group were introduced. Furthermore, style of tai chi included 24-style (3 trials)[, Wustyle (2 trials)[ and Sun-style (1 trial)[, while the control group received treatment including routine care (1 trial)[, no exercise (2 trials)[, health education (1 trial)[, waitlist (2 trials)[, and exercise program (trial=1)[.

3.3 Assessment of risk of bias and quality assessment

Table 2 describes that while 2 RCTs[ mentioned “random” assignment with specific description; most studies were rated as unclear risk (n=2)[ and high risk (n=2)[. Furthermore, the allocation of concealment methods, masked assessors and blinding of the assessments were rated as high risk in all RCTs. Loss to follow-up was described by three RCTs[, which used ITT analysis for incomplete outcome. Furthermore, selective reporting and other bias were rated as high risk and unclear risk in all RCTs, respectively. Furthermore, as only 5 RCTs[ with 6 treatment arms provided meta-analyzable data, we could not conduct a funnel plot analysis to show the presence of risk of publication bias[41].

Table 2.

Evaluation of risk of bias in the six included studies

study	sequence generation	allocation sequence oncealment	blinding of participants and personnel	blinding of outcome assessment	incomplete outcome data	selective outcome reporting	other potential threats to validity
Xie et al 2008[11]	high	high	high	high	low	N/A	low
Gan et al 2007[10]	low	high	high	high	low	N/A	low
Zhou et al 2011[12]	high	high	high	high	low	N/A	low
Ho et al 2012[4]	low	high	high	high	low	N/A	low
Chen et al 2013[13]	N/A	high	high	high	low	N/A	low
Ho et al 2016[5]	N/A	high	high	high	low	N/A	low

N/A=no information available

3.4 Meta-analysis results

3.4.1 Primary outcomes

While there were significant differences between the two groups in the changed total scores of the PANSS (2 trials) (Ho 2016 and Gan 2007), SANS (4 trials) (Xie 2008, Zhou 2011, Ho 2012 and Chen 2013) and SAPS (1 trial) (Chen 2013) over the study period, only 5 RCTs[ with 6 treatment arms provided meta-analyzable data. Compared to control group, we found significantly differences on improvement of negative symptoms (5 trials with 6 treatment arms, n=451, SMD:-0.87 (95% CI: -1.51, -0.24), p=0.007; I2=90%, Figure 2) in the intervention group. Regarding improvement of positive symptoms (4 trials with 5 treatment arms, n=391, SMD:-0.09 (95% CI: -0.44, 0.26), p=0.60; I2=65%, Figure 3), we did not find significant differences between the two groups.

Figure 2.

Adjunctive Tai Chi for schizophrenia: forest plot for improvement of negative symptoms assessed by Positive and Negative Syndrome Scale and Scale for the Assessment of Negative Symptoms

Figure 3.

Adjunctive Tai Chi for schizophrenia: forest plot for improvement of positive symptoms assessed by Positive and Negative Syndrome Scale and Scale for the Assessment of Positive Symptoms

As shown in Table 3, based on the GRADE measure, the strength of the evidence supporting the adjunctive effect of Tai Chi on negative and positive symptoms in schizophrenia was classified as ‘very low’, while discontinuation rate of adjunctive Tai Chi for schizophrenia was classified as “moderate”.

Table 3.

GRADE Analyses: Tai Chi for Schizophrenia

Design	N (arms)	Risk of bias	Inconsistency	Indirectness	Imprecision	Publication bas	Large effect	Overall quality of evidence
Positive symptom score	391(5)	Seriousb	Seriousc	No	No	Seriousd	No	+/-/-/-/; Very Low
Negative symptom score	451(6)	Seriousb	Seriousc	No	No	Seriousd	No	+/-/-/-/; Very Low
Discontinuation rate	334(4)	Seriousb	No	No	No	No	No	+/+/+/-/; Moderate

GRADE = grading of recommendations assessment, development, and evaluation;

aGRADE Working Group grades of evidence: High quality=further research is very unlikely to change our confidence in the estimate of effect. Moderate quality=further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality=further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality=we are very uncertain about the estimate.

bAll studies reported as having a serious bias used aopen label method, only mentioned random allocation without describing the method and withdrawal from the study.

cAll studies reported as having a serious inconsistency had I 50%.

dFor continuous outcomes, N < 400

3.4.2 Secondary outcomes

In the 5 RCTs with meta-analyzable data, social function(3 trials) (Ho 2016, Gan 2007 and Xie 2008), abnormal behaviors(1 trial) (Gan 2007), cognitive function (2 trials)(Ho 2016 and Chen 2013) and perceived stress(1 trial)(Ho 2016) were assessed as secondary outcomes (Table 1). Regarding social function as measured by Inpatient Psychiatric Rehabilitation Outcome Scale (IPROS)[, Social Disability Screening Schedule (SDSS)[ and Treatment Rehabilitation Scale (TRS)[, there was significant improvement in the intervention group. No significant difference was found between the Tai Chi group and the control group in social function as measured by Lawton’s Instrumental Activities of Daily Living (IADL)[ and Activities of Daily Living (ADL)[. In Gan et al’s study[, there was greater improvement in abnormal behaviors as measured by the Nurse’s Observation Scale for Inpatient Evaluation (NOSIE)in the intervention group than the controls. Regarding cognitive function and perceived stress as measured by the Neurological Evaluation Scale (NES)[, backward digit spans of Wechsler Adult Intelligence Scale, Third Edition – Chinese version (WAIS-III)[, Backward Masking Test (BMT) [ and Continuous Performance Test (CPT)[, there was significant improvement in the intervention group. However, there was no significant difference regarding cognitive function assessed the forward digit spans of WAIS-III[ and perceived stress as measured by the Perceived Stress Scale (PSS) [ respectively.

Furthermore, there was no significant difference regarding discontinuation rate between the Tai Chi treatment and the control groups (3 trials with 4 treatment arms, n=334, RR: 0.57 (95% CI: 0.23, 1.40), p=0.22; I2=0%, Figure 4). All included RCTs did not report side effects of Tai Chi.

Figure 4.

Adjunctive Tai Chi for schizophrenia: forest plot for discontinuation rate

3.5 Studies not in the meta-analysis

One study[ were not included in the meta-analysis and found that Tai-Chi could effectively prevented deterioration in movement coordination, interpersonal functioning and disruptions to life activities compared to the control group.

4. Discussion

4.1 Main findings

This systematic review and meta-analysis summarizes results of the effects of Tai Chi exercise on different outcomes in schizophrenia. Evidence from 6 RCTs (n=483) indicated that, Tai Chi as an adjunctive intervention for schizophrenia, appears to have beneficial effects on patients. Specifically, the 5 meta-analyzable RCTs reported that 12-24 weeks of regular Tai Chi activity significantly improved negative symptoms, but had no obvious effects on positive symptoms compared to the controls. In addition, Tai Chi had may improve social function, abnormal behaviors and cognitive function. In contrast, the non-meta-analyzable studies suggested that patients may benefit from Tai Chi in preventing deterioration in movement coordination and interpersonal functioning.

Our results are broadly consistent with recent individual studies and reviews supporting the notion that aerobic exercise can significantly improve psychiatric symptoms[. A recent systematic review[ reported that qigong, another mind body exercise similar to Tai Chi, could be beneficial for improving depressive symptoms. Another meta-analysis of both Chinese and English language publications[ found that Tai Chi was associated with improvements in stress, anxiety, depression and self-esteem.

4.2 Limitation

First, the sample size was relatively small in the 6 RCTs (30 to 153 cases)with limited or incomplete information. Moreover, given the small number of included studies (6 trials), publication bias could not be assessed. This also limited our capability to conduct a more comprehensive data exploration of effect size moderators (i.e. subgroup analysis, sensitivity analysis, and meaningful meta-regression analyses) for improvement of negative symptoms. Second, the 4 RCTs had short treatment duration and hence the effects of Tai Chi’s beyond 24 weeks require further study. Thirdly, prior reviews and meta-analyses indicated Tai Chi has the potential to improving quality of life in other populations[10], therefore such effects for schizophrenia patients need to be further evaluated. Finally, no study provided data on side effects and the quality of this evidence using GRADE was only classified as ‘very low’ (67%) and ‘moderate’ (33%). Thus, high quality studies are needed to confirm and expand on the currently available evidence.

4.3 Implication

Similar to prior meta-analysis[, there is insufficient evidence to determine type- and dose-response effects of Tai Chi on schizophrenia. The studies to date showed a wide variety of types, frequency and duration of Tai Chi. For example, the 24-style was used in three studies, and Wu-style was used in another two. The frequency of Tai Chi exercise varied between 1 and 7 per week, while the duration ranged between 30-60 minutes per session. Furthermore, only short-term effects of Tai Chi were examined in completed studies (12-24 weeks) although the treatment adherence with available data was satisfactory (80-100%). In future studies the intensity, frequency, duration of Tai Chi exercise should be standardized and the dose-response effects need to be further examined.

In traditional Chinese medicine, Tai Chi, which encompasses breathing (Chi) and physical movements, can create physical, emotional and spiritual balance, while enhancing self-fulfillment and self-realization[. To date, however, the underlying mechanism of action of Tai Chi’s effects on psychiatric disorders including schizophrenia has not been identified. The possible mechanisms of Tai Chi exercise for enhanced psychiatric symptoms and related outcomes may result from its beneficial influence on physiological, neurological, cardiovascular and immunological effects and overall well-being[. In addition, it should be noted that all included studies were conducted in Chinese treatment settings where cultural beliefs may lead to an expectancy of benefits from Tai Chi. Therefore effects of Tai Chi on schizophrenia need to be replicated in non-Chinese settings in the future.

This is the first meta-analysis of Tai Chi as an adjunctive treatment for patients with schizophrenia. Most previous meta-analyses and reviews have focused on the effect of Tai Chi on psychological well-being in major depressive disorder[. Tai chi is an inexpensive and safe mind-body practice and may enhance the general well-being of schizophrenia patients, therefore could be integrated into the complex management of schizophrenia. The major strength of this study is that the meta-analysis was based on recent, methodically rigorous RCTs published in both English and Chinese languages. This meta-analysis gave the international readership access to studies on Tai Chi published in Chinese journals, which are rarely available. Tai Chi is popular and widely practiced now in the general population all over the world therefore finding qualified Tai Chi practitioners to employ by mental health services would not be difficult.