Table 2.

Reported clinical trials of Tai Chi for patients with Parkinson disease

Study	Size	Design	Duration	Intervention	Controls	Outcomes	AEs
Venglar (2005) [41]
USA	1	Open-label	8 weeks	Weekly hour-long Tai Chi class	–	Motor: improvement in TUG test—both speed and number of steps Qualitative: increased awareness of how patient was moving, worsening of balance confidence on ABC scale	Did not report
Li et al. (2007) [42]
USA	17	Open-label	5 days	Daily 90-min Tai Chi exercise program	–	Qualitative: The program was well received by all participants with respect to program appropriateness, participant satisfaction and enjoyment, and intentions to continue	The program was safe. No serious adverse events
Kim et al. (2014) [40]
Korea	12	Open-label	12 weeks	3 Yang-style Tai Chi sessions per week	–	Motor: statistically significant improvement in anteroposterior (124%) and mediolateral (135%) displacement of the center of pressure compared to baseline	Did not report
Hackney and Earhart (2008) [43]
USA	33	Randomized, controlled	10–13 weeks	20 60-min Yang-style Tai Chi training sessions (n = 17)	Non-exercise control group (n = 16)	Motor: significant improvement on BBS in the Tai Chi group relative to controls; non-significant trends toward improvement on mUPDRS, 6MWT, TUG test, tandem stance, and backward walking compared to controls Qualitative: All Tai Chi participants reported that they enjoyed the program and generally noted improvement in subjective measures of motor function	Did not report
Hackney and Earhart (2009) [44]
USA	61	Randomized, controlled	20 weeks	40 sessions of Tai Chi (n = 13), tango (n = 14), waltz/foxtrot (n = 17)	Non-exercise control group (n = 17)	Non-motor: The tango group had significant improvement in PDQ-39 summary index, social support, and mobility subscales	Did not report
Amano et al. (2013) [45]
USA	45	Randomized, controlled	16 weeks	2 or 3 60-min Yang-style Tai Chi sessions per week (n = 27)	Qi Gong meditation (n = 9) or usual care (n = 9)	Motor: ineffective in improving gait initiation, gait performance, or mUPDRS compared to controls	Did not report
Nocera et al. (2013) [46]
USA	21	Randomized, controlled	16 weeks	3 60-min Yang-style Tai Chi sessions per week (n = 15)	Non-exercise control group (n = 6)	Non-motor: significant improvement on PDQ-39 total score (effect size = 1.03, p = 0.04) and emotional well-being subscale (effect size = 0.46, p = 0.04)	Did not report
Choi et al. (2013) [47]; Choi (2016) [48]
Korea	20	Randomized, controlled	12 weeks	3 60-min Tai Chi sessions per week (n = 11)	Non-exercise control group (n = 9)	Motor: significant improvement from baseline on the mUPDRS in the Tai Chi group, and interaction effects (time vs group) on reaction time and one leg balance; a secondary analysis of functional fitness showed significant improvement in arm curls, functional reach, and one leg balance compared to baseline. Non-motor: significant improvement from baseline on UPDRS mentation, behavior, and mood subscale	Did not report
Gao et al. (2014) [49]
China	76	Randomized, controlled	12 weeks	3 60-min Yang-style Tai Chi sessions per week (n = 37)	Non-exercise control group (n = 39)	Motor: significant improvement on the BBS compared to controls, though no improvement in mUPDRS or TUG test; the Tai Chi group had fewer falls than controls	Did not report
Vergara-Diaz et al. (2018) [50]
USA	32	Randomized, wait-list-controlled	6 months	Average of 2 Tai Chi sessions per week (n = 16)	Non-exercise control group (n = 16)	Motor: non-significant trend toward improvement in ABC and TUG test Non-motor: non-significant trend toward improvement in PDQ-39; both groups improved on trail-making test, though not significant	No adverse events related to the intervention
Li et al. (2012) [51]; Li et al. (2014) [52]
USA	195	Randomized, controlled	24 weeks	2 60-min Tai Chi sessions per week (n = 65)	Resistance training (n = 65), stretching (n = 65)	Motor: Tai Chi did better than the resistance training and stretching group in maximum excursion and directional control (p < 0.001 for both). The Tai Chi group performed significantly better than stretching on all measures and outperformed the resistance training group on stride length and functional reach. Tai Chi lowered the incidence of falls as compared with stretching but not resistance training. Non-motor: significant improvement in PDQ-8 compared to resistance training, and both PDQ-8 and VPS compared to the stretching group	No serious adverse events
Cheon et al. (2013) [53]
Korea	23	Randomized, controlled	8 weeks	3 60-min Sun-style Tai Chi sessions per week (n = 9)	Combined exercise (n = 7), non-exercise controls (n = 7)	Motor: non-significant trend toward improvement in UPDRS in both exercise groups; the combined exercise group performed better on measures of strength, flexibility of the upper limbs, and cardiovascular endurance, while the Tai Chi group performed better for measures of flexibility, strength of the lower limbs, agility, and cardiovascular endurance. Non-motor: tendency for improvement in quality of life scales in both exercise groups: emotional QOL in the Tai Chi group and social QOL in the exercise group	No serious adverse events
Zhang et al. (2015) [54]
China	40	Randomized, controlled	12 weeks	2 60-min Yang-style Tai Chi sessions per week (n = 20)	Multimodal exercise training (n = 20)	Motor: no between-group difference in the BBS; improvement from baseline in both groups in mUPDRS, stride length, gait velocity, TUG test, though no between-group differences Qualitative: Tai Chi was harder to incorporate into a daily routine than exercise training.	No serious adverse events in either exercise group
Yang et al. (2017) [55]
China	36	Randomized, controlled	13 weeks	3 60-min Yang-style Tai Chi sessions per week (n = 19)	Individual-based Yang-style Tai Chi (n = 17)	Motor: No statistically significant differences between groups Non-motor: Both groups had improvement from baseline in global non-motor symptoms and sleep; group-based had better cognition on MoCA (p = 0.002); no improvement in either group on HAMD Qualitative: Group-based had higher home exercise compliance rate (64.84% vs 51.17%, p = 0.019)	Did not report
Kurt et al. (2018) [56]
Turkey	40	Randomized, controlled	5 weeks	5 60-min Ai Chi sessions per week (n = 20)	5 60-min land-based exercises per week (n = 20)	Motor: significantly improved dynamic balance (p < 0.001), BBS (p < 0.001), TUG test (p = 0.002), and mUPDRS (p < 0.001) in the Ai Chi group compared with controls. Both groups did improve relative to baseline in all measures. Non-motor: PDQ-39 was significantly improved from baseline in both groups. The Ai Chi group had significant improvement compared with the land-based exercise group	Did not report

6MWT = 6-min walk test; ABC = Activity-Specific Balance Confidence; AEs = adverse events; BBS = Berg Balance Scale; HAMD = Hamilton Depression Rating Scale; MoCA = Montreal Cognitive Assessment; mUPDRS = Unified Parkinson’s Disease Rating Scale, motor subscale; PDQ-8 = 8-item Parkinson’s Disease Questionnaire; PDQ-39 = 39-item Parkinson’s Disease Questionnaire; QOL = quality of life; TUG = Timed Up and Go; UPDRS = Unified Parkinson’s Disease Rating Scale; VPS = Vitality Plus Scale