The Effect of Tai Chi on Health Outcomes in Patients With Chronic Conditions: A Systematic Review

Objective  To conduct a systematic review of reports on the physical and psychological effects of Tai Chi on various chronic medical conditions.

Data Sources  Search of 11 computerized English and Chinese databases.

Study Selection  Randomized controlled trials, nonrandomized controlled studies, and observational studies published in English or Chinese.

Data Extraction  Data were extracted for the study objective, population characteristics, study setting, type of Tai Chi intervention, study design, outcome assessment, duration of follow-up, and key results.

Data Synthesis  There were 9 randomized controlled trials, 23 nonrandomized controlled studies, and 15 observational studies in this review. Benefits were reported in balance and strength, cardiovascular and respiratory function, flexibility, immune system, symptoms of arthritis, muscular strength, and psychological effects.

Conclusions  Tai Chi appears to have physiological and psychosocial benefits and also appears to be safe and effective in promoting balance control, flexibility, and cardiovascular fitness in older patients with chronic conditions. However, limitations or biases exist in most studies, and it is difficult to draw firm conclusions about the benefits reported. Most indications in which Tai Chi was applied lack a theoretical foundation concerning the mechanism of benefit. Well-designed studies are needed.

Tai Chi is a traditional Chinese martial art that has been practiced in China for many centuries. It combines deep diaphragmatic breathing and relaxation with many fundamental postures that flow imperceptibly and smoothly from one to the other through slow, gentle, graceful movements.^1-5 It has been advocated for development of mind-body interaction, breathing regulation with body movement, hand-eye coordination, and tranquilization.^6-11

Tai Chi has evolved into many different styles during its development, including Chen, Wu, Sun, and Yang style. Among these styles, Chen is the oldest, while Yang is the most popular. Despite the lack of rigorous evidence regarding its benefits, Tai Chi is widely practiced in many countries as a form of exercise for health and fitness.

Tai Chi is practiced as an exercise to promote good health, memory, concentration, digestion, balance, and flexibility and is also thought to improve psychological conditions such as anxiety, depression, and declines associated with aging and inactivity. It is also practiced to improve quality of life.^9-11 However, despite its popularity, the biological mechanism and clinical effects of Tai Chi are not well understood. The purpose of this systematic review is to summarize the studies that have examined the effect of Tai Chi on patients with a variety of chronic conditions and to identify and describe the limitations and biases of these published clinical studies.

A total of 743 abstracts pertaining to the practice of Tai Chi were obtained from 11 sources: (1) MEDLINE search from 1966 through April 2002 (87 abstracts); (2) SPORTDiscus Database search from 1949 to December 2000 (419 abstracts); (3) Social Sciences Abstracts search from 1984 to December 2000 (19 abstracts); (4) Health Star search from 1975 to 2000 (22 abstracts); (5) PsycINFO search from 1887 to May 2001 (39 abstracts); (6) ERIC search from 1966 to May 2001 (13 abstracts); (7) AIDSLINE search from 1980 to December 2000 (4 abstracts); (8) Biological Abstracts search from 1980 to March 2001 (30 abstracts); (9) Sociological Abstracts search from 1963 to December 2000 (3 abstracts); (10) CINAHL search from 1982 to May 2001 (28 abstracts); and (11) Chinese Medical Database from 1976 to June 2000 (obtained from Nanjing Medical University Library) (79 abstracts). The search strategies used the text word "Tai Chi." This includes English and Chinese articles and all types of studies.

Original studies were included in the systemic review if they evaluated Tai Chi for treatment of a chronic condition and reported outcome data. Randomized controlled trials (RCTs), nonrandomized controlled studies (NRSs), and observational studies qualified.

The following criteria were used to evaluate study quality: (1) well-defined study question; (2) explicit and/or appropriate eligibility criteria; (3) proper allocation of intervention groups; (4) use of groups with similar baseline characteristics; (5) confounders accounted for; (6) interventions and outcomes adequately described; (7) blinded outcomes assessment; (8) valid outcome measurements and statistical methods; (9) adequate follow-up rate; dropout rate reported; and (10) conclusions supported by the findings. Two investigators extracted data.

Articles were categorized into clinical domains. For each clinical domain, we summarized information from each study. The summary tables described the interventions of Tai Chi, the outcomes measured, and the authors' main conclusions when appropriate. Included studies for each clinical domain were also assessed to determine the strengths and limitations of the most important studies following a detailed rationale for the appraisal of study characteristics related to quality. Because of the heterogeneity of outcomes, study designs, and settings, we did not perform a meta-analysis on the outcomes.

All abstracts were reviewed to identify the relevant studies related to the effects of Tai Chi. Of the 743 abstracts initially identified, 679 were excluded because they were review articles, case reports, letters or comments, conference proceedings, information about Tai Chi classes or training programs, theses or dissertations, newspaper articles, announcements, or duplicate publications.

After the initial screening, 64 articles were retrieved and reviewed. Seventeen studies were eliminated because they were duplicate publications, English translations of original Chinese articles, or contained major methodologic flaws such as study populations that did not meet the eligibility criteria, measurement tools that lacked rigorous validity and had not been tested with the population under consideration, and inadequately reported outcomes.

Ultimately, 47 studies (9 RCTs, 23 NRSs, and 15 observational studies) related to the clinical issues were identified for data abstraction and critical appraisal (Table 1). The characteristics of the original research articles were assessed for each of 7 conditions.^12-57 Our reported results correspond to the data on each condition and are summarized below.

The evidence on balance comprised 11 studies^12-23 (Table 2): 2 RCTs of 24 and 200 subjects, respectively; 5 NRSs of 125 subjects in total; 3 cross-sectional studies with a total of 104 subjects; and 1 follow-up study of 110 subjects. Balance control, maximal voluntary extension, strength, flexibility, cardiovascular endurance, and postural stability were measured in these studies.

Seven clinical trials^12-19 (2 RCTs and 5 NRSs) reported that 8 to 16 weeks of Tai Chi training significantly improved balance, flexibility, and strength of knee extension and reduced the occurrence of falling in community-dwelling elders. One follow-up study²⁰ of 110 community-dwelling persons (mean age, 80 years) that combined strengthening and weight-training programs for 3 months, then 6 months of Tai Chi training, found a favorable impact on a variety of balance measures. Three cross-sectional studies^21-23 of individuals with 1 to 35 years of Tai Chi practice experience revealed that long-term Tai Chi practitioners had greater lower extremity flexibility than nonpractitioners and that Tai Chi was effective in improving normal gait velocity and producing a trend toward improved maximal gait velocity in elders. Overall, these studies reported that long-term Tai Chi practice had favorable effects on the promotion of balance control, flexibility, and cardiovascular fitness and reduced the risk of falls in elders.

However, most studies were NRSs,^15-19 had no comparison group,^15-17 or had relatively small sample sizes.^14-17,21,22 Other studies lacked detailed information on health status and eligibility criteria,^15-17 and blinding assessment of outcomes were not well reported in some studies.^17,19,20,23 In others, potential selection bias^15-17 and uncontrolled confounding factors such as age, body mass index, sex, exercise time, and a large discrepancy in Tai Chi practitioner experience might also have existed.^14,20-23 Furthermore, the cross-sectional studies were too limited to explain the cause-effect relationships.^21-23 The differences between styles of Tai Chi exercise, especially between the traditional Chinese styles (Chen, Yang, and Wu) and self-modified as well as westernized Tai Chi forms, were not made clear. The personalities of participants and the abilities of different Tai Chi instructors may vary. Several studies reported a wide range of Tai Chi exercise experience, spanning from 1 year to 35 years, making it difficult to relate the amount of benefit to the duration of exercises.^21-23

Four studies^24-27 evaluated Tai Chi effects on musculoskeletal conditions (Table 3). One RCT²⁴ of 33 patients with osteoarthritis reported that 12 weeks of Tai Chi practice significantly improved arthritis symptoms, self-efficacy, level of tension, and satisfaction with general health status. Functional capacity (1-leg standing balance, 50-foot (15-m) walking speed, and time to rise from a chair), arthritis self-efficacy, and quality of life (Arthritis Impact Measurement Scale) were measured.

One publication²⁵ reported 2 NRSs of 47 and 28 patients with rheumatoid arthritis, respectively, who underwent 10 weeks of Tai Chi training. Disease activity (joint tenderness and number of swollen joints) and exacerbation of joint symptoms were measured; 50-foot (15-m) walk and handgrip strength were measured; and a written functional assessment was performed. The study showed no significant differences between the Tai Chi and control groups in disease activity. The authors suggested that Tai Chi appeared to be safe for patients with rheumatoid arthritis and might serve as a weight-bearing exercise with additional potential advantages of stimulating bone growth and strengthening connective tissue.

Another NRS²⁶ described 41 community-dwelling subjects (mean age, 62 years) who participated in a 6-month Tai Chi training course. The researchers found that concentric knee extensor peak torque increased by 15% to 20%, and eccentric peak torque increased by 15% to 24% in men. The women also showed increases, ranging from 14% to 22% in concentric peak torque and 18% to 24% in eccentric peak torque. In addition, the knee extensor endurance ratio increased by 10% to 19% in men and 10% to 15% in women. Therefore, the study concluded that Tai Chi training may enhance muscular strength and endurance of knee extensors in elderly individuals.

An NRS study²⁷ of 19 patients with multiple sclerosis involved an 8-week Tai Chi training course that measured walking speed, hamstring flexibility, and psychosocial well-being using the Medical Outcomes Study 36-Item Short-Form General Health Survey. The results revealed that Tai Chi increased walking speed and hamstring flexibility. Patients experienced improvements in vitality, social functioning, mental health, and ability to carry out physical activities and emotional roles. The study concluded that Tai Chi maximized independence and improved quality of life for people with chronic disabling conditions.

The limitations of these studies included small sample size,^24-27 lack of randomization,^25-27 loss to follow-up,^25,26 potential selection bias, uncontrolled confounding factors,^25-27 unclear statistical analysis,^25-27 lack of standardized outcome measures,^25-27 and lack of blinded assessment of outcome.^25-27

We found 2 RCTs,^28,29 and 2 NRSs^30,31 with a total of 401 patients with hypertension (Table 4). The duration of Tai Chi training for the studies was 8 to 12 weeks^28,29,31 and 3 years.³⁰ Blood pressure, maximal oxygen uptake, and heart rate (HR) were recorded before and after each session. A reduction of mean blood pressure was found for regular Tai Chi practice in all the studies. Young et al²⁸ found adjusted mean (SE) changes in systolic blood pressure during the 12-week intervention period of –8.4 (1.6) mm Hg and –7.0 (1.6) mm Hg in the aerobic exercise and Tai Chi groups, respectively (within-group P<.001; between-group P = .56). Corresponding changes for diastolic blood pressure were –3.2 (1.0) mm Hg in the aerobic exercise group and –2.4 (1.0) mm Hg in the Tai Chi group (within-group P<.001; between-group P = .54). Channer et al²⁹ reported that over 11 sessions of exercise, both aerobic and Tai Chi exercise were associated with reduction in systolic blood pressure (P<.05), and Tai Chi was also associated with a reduction in diastolic blood pressure (P<.01) in patients recovering from acute myocardial infarction. Similar findings were also reported in 2 NRSs^30,31 conducted in China.

These studies included biases due to volunteer effect, confounding factors, and loss to follow-up.^28,30,31 The eligibility criteria were not clearly specified,^29-31 and several studies lacked standardized outcome measurements and appropriate statistical analysis.^30,31

Numerous studies have evaluated the effects of Tai Chi on cardiovascular and respiratory function.^32-48 Since 1993, three observational studies and 2 NRSs^32-36 were conducted by Taiwan groups to evaluate the cardiovascular effect of regular Tai Chi practice (Table 5). Two cohort studies^32,33 of 90 and 84 subjects with 6.3 and 6.7 years' experience of Tai Chi, respectively, were reported. Cardiopulmonary exercise testing using incremental cycle ergometry and HR measurements were performed in these studies. The researchers found that the oxygen uptake (V̇O₂ and work rate in the Tai Chi group were significantly higher than in the control group. Tai Chi practitioners had a smaller decline in the maximum oxygen uptake than their sedentary counterparts.

One case control study³⁴ of 76 subjects with 11.8 years of Tai Chi practice (mean age, 69 years) evaluated cardiovascular function, flexibility, and body composition. The researchers found that the long-term Tai Chi practitioners showed higher V̇O₂ scores in the stand-and-reach test and a lower percentage of body fat than their sedentary counterparts. Two NRSs^35,36 evaluated the training effect of Tai Chi for 38 community-dwelling subjects (aged 58-70 years) and 20 low-risk patients who underwent coronary artery bypass surgery (mean age, 57 years). After 1 year of 4-times-weekly Tai Chi training, the Tai Chi group showed significantly enhanced cardiorespiratory function, strength and flexibility, and increased V̇O₂ and work rate compared with the control group who performed self-adjusted exercises.

One NRS³⁷ with 20 subjects (aged 30-35 years) examined the metabolic and cardiorespiratory responses to continuous performance of Tai Chi and Wing Chun (rapid forceful striking and kicking movements with forced expirations timed with punching and kicking in an attempt to achieve maximum power). The exercise values corresponded to 52.4% of maximum oxygen uptake values (VO_2max) and 70.5% of maximum HR (HR_max) for Wing Chun and only 36.4% of VO_2max and 59.8% of HR_max for Tai Chi. The ventilatory equivalent for V̇O₂ obtained during Tai Chi (21.7) was significantly lower than for Wing Chun exercise (24.2). The authors concluded that Tai Chi practitioners use efficient breathing patterns during exercise. However, because Tai Chi is a low- to moderate-intensity form of exercise, it might not be suitable for improving cardiorespiratory fitness.

Results related to the effect of Tai Chi on cardiovascular and pulmonary function have been reported in 11 publications by Chinese researchers since 1979.^38-48 Although 1 study⁴⁰ reported that the metabolic intensity of the activity seemed insufficient to generate improvements of cardiorespiratory fitness in healthy young adults, all other studies suggested that regularly practiced Tai Chi might delay the decline of cardiorespiratory function in older individuals and might be prescribed as a suitable exercise for older adults.

No RCTs were identified in this area, and the overall study quality was very poor, especially among the studies conducted in China. Many limitations were found in the Chinese studies, such as unclear study designs,^32,34,36-48 small sample sizes,^{36-38,40,41,43-45,47} poorly reported baseline health status and eligibility criteria of the subjects,^38-48 lack of comparison groups,^41,45-47 lack of detailed information about the types of Tai Chi exercise and duration of the trial,^{37,39,40,42,43,46-48} lack of accounting of subjects lost to follow-up,^38-48 and no blinding of outcome assessors.^38-48 In addition, study limitations included potential selection bias,^32,33,37-48 confounding factors,^32,33,37-48 inconsistent measurement intervals,^{33-35,39-42,44-48} and large differences in Tai Chi practitioner experience.^{33-35,39-42,44-48}

The effect of Tai Chi on psychological responses were evaluated in 3 RCTs^49-51 and 3 NRSs^52-54 (Table 6). Results from 2 RCTs indicated that 283 low-activity older adults participating in either a 16-week or a 6-month Tai Chi exercise program showed improvement compared with the control groups in several indices of psychological well-being that evaluated depression, psychological distress, positive well-being, life satisfaction, and perceptions of health.^49,50

Another RCT⁵¹ examined the psychological effect of a 12-week Tai Chi program on 90 schoolchildren, grades 4 to 6 (52 boys, 38 girls), who were pretested on measures of perceived self-competence, visual-motor integration, and anxiety. The Tai Chi group significantly improved their scores on the perceived self-competence and visual-motor integration tests.

Two NRSs^52,53 with a total of 186 patients reported that 1 to 46 months of Tai Chi practice improved mood and reduced stress and anxiety. Nine elderly patients diagnosed with multiple-infarct dementia or Alzheimer disease participated twice weekly over 7 weeks in a before-and-after Tai Chi trial.⁵⁴ The authors concluded that "structured reminiscence with Tai Chi facilitated thinking that was focused and insightful, beyond the level normally manifested for this group of participants."

Although these 6 studies (including 3 RCTs) were conducted in 4 countries (United States, Canada, China, and Australia), the study populations were poorly defined. The Tai Chi interventions were not adequately described, and there was no blinding of outcome assessors to intervention in any of these studies.

Two studies^55,56 evaluated the effects of Tai Chi practice on the endocrine or immune systems (Table 7). A cross-sectional Chinese study⁵⁵ of 98 elderly men reported that 10 years of Tai Chi practice might widely affect endocrine function, including the pituitary-thyroid system and the pituitary-gonad system, and may strengthen pituitary metabolic reaction among elderly men. The conclusion of this study is questionable because cross-sectional studies are not designed to evaluate causality.

An NRS of 60 elderly subjects⁵⁶ found that the total number of circulating T cells, including active T lymphocytes, were significantly higher in the Tai Chi group (30 healthy subjects, aged ≥60 years, who regularly practiced Tai Chi for 4 or more years) than in the untrained group (30 age-matched subjects).

Both studies lacked explicit and appropriate eligibility criteria. In both studies, the Tai Chi intervention was not well described, and the follow-up rate and statistical analysis were not adequately reported.

Several studies evaluated the beneficial effects of Tai Chi in other areas (Table 8).^57-59 A recent RCT⁵⁷ found that a 60-minute Tai Chi practice session twice a week for 6 months using a classical Yang style could significantly enhance self-efficacy in older adults. Two aspects of self-efficacy were measured: barriers and performance. The study also suggested that changes in self-efficacy cognitions were significantly related to class attendance.

A recent case-control study⁵⁸ of 20 elderly men revealed that 10 elderly men who practiced Tai Chi for 11.2 years had a 34% higher V̇O₂ peak and higher skin blood flow, cutaneous vascular conductance, and skin temperature than the 10 sedentary men at rest and during exercise. Another NRS⁵⁹ of 22 young people (mean age, 21 years) found that 20 days of Tai Chi practice significantly decreased nightmares.

Most of the studies evaluated in this systematic review have been performed in China and the United States to examine the physiological and psychosocial benefits of Tai Chi for all age groups. Benefits were reported by the authors of these studies in cardiovascular and respiratory function in healthy subjects and in patients who had undergone coronary artery bypass surgery as well as in patients with heart failure, hypertension, acute myocardial infarction, arthritis, and multiple sclerosis. Benefit was also found for balance, strength, and flexibility in older subjects; falls in frail elderly subjects; and pain, stress, and anxiety in healthy subjects. Overall, Tai Chi appears to have physiologic and psychosocial benefits and appears to be safe and effective in promoting balance control, flexibility, and cardiovascular fitness for older adults with chronic conditions.

Unfortunately, many studies of Tai Chi lack rigorous scientific methods, and most investigations have been retrospective and have not used randomized control groups. It is also difficult to obtain overall quantitative estimates of treatment effects owing to the heterogeneity of inclusion criteria, patients, and outcome definitions and inadequate information on design, details of the intervention, and outcomes. It should also be noted that none of the studies from Asia were RCTs. In contrast to those published in the United States and other Western countries, almost all the studies published in mainland China, Hong Kong, and Taiwan reported positive results. Studies may have been conducted with different levels of methodologic rigor, and publication bias may be greater in some countries than in others.⁶⁰

In addition, the mechanisms of the benefits from practicing Tai Chi for any of the conditions studied are not well understood. There are only 9 RCTs in this review, and they examined only short-duration practice of Tai Chi (8-16 weeks). Therefore, long-term effects of Tai Chi practice are still unknown, and there is insufficient information to recommend Tai Chi to patients with chronic conditions. Well-defined study questions, adequate selection criteria, groups similar at baseline, valid statistical methods, accounted-for confounders, appropriate outcome, and adequate follow-up are needed for proper evaluation of the effects of Tai Chi. Patients and physicians who use Tai Chi intervention will be better informed by high-quality RCTs that report short- and long-term risks and benefits.

Corresponding author and reprints: Chenchen Wang, MD, MSc, Institute for Clinical Research and Health Policy Studies, Tufts-New England Medical Center, 750 Washington St, No 63, Boston, MA 02111 (e-mail: Cwang@tufts-nemc.org).

Accepted for publication March 28, 2003.

This work was partially supported by the General Clinical Research Center, funded by the Division of Research Resources of the National Institutes of Health under grant No. M01-RR0054, and by funding from the Tufts-New England Medical Center (Dr Wang).