Abstract
OBJECTIVES: This systematic review aims to summarize the existing literature on Tai Chi randomized controlled trials (RCTs) and recommend Tai Chi exercise prescriptions for different diseases and populations. METHODS: A systematic search for Tai Chi RCTs was conducted in five electronic databases (PubMed, Cochrane Library, EMBASE, EBSCO, and Web of Science) from their inception to December 2019. SPSS 20.0 software and Microsoft Excel 2019 were used to analyze the data, and the risk of bias tool in the RevMan 5.3.5 software was used to evaluate the methodological quality of RCTs. RESULTS: A total of 139 articles were identified, including diseased populations (95, 68.3%) and healthy populations (44, 31.7%). The diseased populations included the following 10 disease types: musculoskeletal system or connective tissue diseases (34.7%), circulatory system diseases (23.2%), mental and behavioral disorders (12.6%), nervous system diseases (11.6%), respiratory system diseases (6.3%), endocrine, nutritional or metabolic diseases (5.3%), neoplasms (3.2%), injury, poisoning and certain other consequences of external causes (1.1%), genitourinary system diseases (1.1%), and diseases of the eye and adnexa (1.1%). Tai Chi exercise prescription was generally classified as moderate intensity. The most commonly applied Tai Chi style was Yang style (92, 66.2%), and the most frequently specified Tai Chi form was simplified 24-form Tai Chi (43, 30.9%). 12 weeks and 24 weeks, 2-3 times a week, and 60 min each time was the most commonly used cycle, frequency, and time of exercise in Tai Chi exercise prescriptions. CONCLUSIONS: We recommend the more commonly used Tai Chi exercise prescriptions for different diseases and populations based on clinical evidence of Tai Chi. Further clinical research on Tai Chi should be combined with principles of exercise prescription to conduct large-sample epidemiological studies and long-term prospective follow-up studies to provide more substantive clinical evidence for Tai Chi exercise prescriptions.