This site is an archive of a closed Robert Wood Johnson Foundation program, provided for educational and historical purposes. Please note that this content is not routinely updated and that contact information and social links may not work.

The Safety and Feasibility of a Tai Chi Intervention for Stroke Survivors: A Pilot Study

Objective: Examine the safety and feasibility of a 12-week Tai Chi intervention among stroke survivors.

Design: Two-group, prospective pilot study with random allocation.

Setting: Outpatient rehabilitation facility.

Subjects: Stroke survivors ≥50 years and at ≥three months post-stroke.

Interventions: Tai Chi subjects attended group-based Yang Style classes three times/week for 12-weeks, while Usual Care subjects received weekly phone calls along with written materials/resources for participating in community-based physical activity.

Main outcome measures: Indicators of study safety and feasibility included recruitment rates, intervention adherence, falls or adverse events, study satisfaction, drop-outs, and adequacy of the outcomes measures.

Results: Interested persons pre-screened by phone (n = 69) were on average 68 years old, (SD = 13) years old, 48% (n = 33) women, 94% (n = 65) were at least three months post-stroke. A total of 28 subjects aged 69 (SD = 11) years enrolled in this pilot study. Intervention adherence rates were very high (≥92%). There were no falls or other adverse events. The dose of Tai Chi exercise (≥150 minutes/week) was well tolerated. Overall study satisfaction was high (8.3 (SD = 1.9); 1 = not satisfied, 10 = most satisfied), while drop-outs (n = 3, 11%) were unrelated to study intervention. Score distributions for the outcome measures were approximately normal, sensitive to change, and seemed to favor the Tai Chi intervention.

Conclusions: Tai Chi is a safe, community-based exercise program for stroke survivors. Our data suggest that recruitment and retention of an adequate sample is feasible, and that in a full-scale study 52 subjects/group are needed to detect statistically significant between group differences (alpha = 0.05, power = 0.80).