Table 4.

Reported clinical trials of yoga for patients with Parkinson disease

Study	Size	Design	Duration	Intervention	Controls	Outcomes	AEs
Hall et al. (2011) [68]
UK	1	Case report	8 weeks	Weekly 60-min yoga session	–	Motor: The participant had a non-significant improvement on BBS and TUG test. There was no consistent trend in hip and ankle range of motion. Non-motor: no change from baseline in the PDQ-39	Did not report
Moriello et al. (2013) [69]
USA	1	Case report	6 months	Twice weekly 90-min yoga session plus a home exercise program for 12 weeks, followed by a personalized home exercise program incorporating yoga	–	Motor: The participant had an improvement from baseline to 6 months on the HiMAT (+ 11 points). Non-motor: PDQ-39 score improved after 6 months (− 16 points)	Did not report
Boulgarides et al. (2014) [70]
USA	10	Open-label	8 weeks	Weekly 60-min yoga session	–	Motor: no change in mUPDRS, BBS, mDGI, and FRT after the intervention; near-significant improvement in lower extremity strength and flexibility, single limb balance, and 30-s chair stand after the intervention Non-motor: near-significant improvement on HADS after the intervention Qualitative: 5/10 continued to participate in paid yoga classes	No adverse events
Colgrove and Sharma (2012) [71]; Sharma et al. (2015) [72]
USA	13	Randomized, controlled	12 weeks	Twice weekly 60-min Iyengar Hatha yoga sessions (n = 8)	Usual care (n = 5)	Motor: significant improvement in mUPDRS compared to baseline (p = 0.006), though there was no difference between groups; the yoga group also had improvement in diastolic BP (p = 0.036) and average forced vital capacity (p = 0.03). Non-motor: trend toward significance on the GDS after intervention Qualitative: Yoga participants reported more positive symptom changes including tremor reduction, more relaxation, and less fatigue; 2 patients reduced medications	No adverse events
Cheung et al. (2018) [73]
Australia	20	Randomized, wait-list-controlled	12 weeks	Twice weekly 60-min group-based Hatha yoga (n = 10)	Usual care (n = 10)	Motor: The mUPDRS was improved compared to controls after the intervention. Non-motor: no change in sleep quality, depression, cognitive function, and global quality of life; the yoga group had worse scores on the sleep and outlook subscales on PDQUALIF. Qualitative: Most (17/20) stated they “definitely enjoyed” the program	No adverse events
Van Puymbroeck et al. (2018) [74]; Walter et al. (2019) [75]
USA	27	Randomized, wait-list-controlled	8 weeks	Twice weekly 60-min Hatha yoga sessions (n = 15)	Usual care (n = 12)	Motor: greater functional gait in the yoga group compared to controls; improvement in mUPDRS and freezing of gait compared to baseline; within the yoga group, significant improvement from baseline in postural stability (mini-BESTest), fall risk, Activity-Specific Balance Confidence Scale, Falls Management Scale Non-motor: Within the yoga group, there was a significant improvement in PFS-16 and PDQ-8 from baseline	Did not report
Bega and Stein (2016) [76]
USA	17	Randomized, single-blind, controlled	12 weeks	Twice weekly 60-min Iyengar yoga (n = 9)	Resistance training (n = 8)	Motor: Both groups had improvement from baseline on the mUPDRS, TUG test, and BBS, but no significant between-group differences. Non-motor: Both groups had improvement from baseline on PDQ-39, but no significant between-group difference. Qualitative: The program was feasible, as 13/17 went to at least 75% of classes, though this was significantly higher than the resistance training group. In the yoga group, 8/9 agree that class was enjoyable and beneficial	No adverse events
Ni et al. (2016a) [77]; Ni et al. (2016b) [78]
USA	41	Randomized, controlled	12 weeks	Twice weekly 60-min group power yoga sessions (n = 15)	Twice weekly power training (n = 14), non-exercise controls participating in a 1-h per month health education course (n = 12)	Motor: Both exercise groups scored significantly better on mUPDRS, BBS, mini-BESTest, TUG test, functional reach (less affected side), 10-m walking, muscle strength and power (1 repetition maximum), and peak power compared to controls, but there were no between-group differences. In a secondary analysis evaluating yoga versus non-exercise controls, the yoga group had significant improvement from baseline in bradykinesia, rigidity, muscle strength, and power. Non-motor: There was significant improvement on the PDQ-39 sum compared to controls, and in the subdomains of mobility and ADLs.
Kwok et al. (2019) [79]
China	138	Randomized, single-blind, controlled	8 weeks	Weekly 90-min Hatha yoga sessions (n = 71)	Weekly 60-min stretching and resistance training exercise (n = 67)	Motor: Both groups had improvement in mUPDRS and TUG test compared to baseline, but there were no between-group differences. Non-motor: There was a significant improvement in HADS compared to controls

ADLs = activities of daily living; AEs = adverse event; BBS = Berg Balance Scale; BP = blood pressure; FRT = Functional Reach Test; GDS = Geriatric Depression Scale; HADS = Hospital Anxiety and Depression Scale; HiMAT = High-level Mobility Assessment Tool; mDGI = modified Dynamic Gait Index; Mini-BESTest = Mini Balance Evaluation Systems Test; 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; PDQUALIF = Parkinson’s Disease Quality of Life Scale; PFS-16 = Parkinson’s Disease Fatigue Scale; TUG = Timed Up and Go