Long-term effects of highly challenging balance training in Parkinson's disease-a randomized controlled trial

Objectives: To determine long-term effects of a highly challenging training program in people with Parkinson's disease, as well as describe how initially observed improvements of the program deteriorated over time. Design: Long-term follow-up of previously reported outcomes at 10 weeks of a randomized controlled trial. Setting: University hospital setting. Participants: One-hundred elderly with mild-to-moderate (Hoehn and Yahr 2-3) Parkinson's disease. Interventions: Participants in the training group (n = 51) received 10 weeks (three times/week) of balance and gait exercises, incorporating dual-tasks, while the control group (n = 49) received care as usual. Main Outcome Measures: Balance control (Mini-Balance Evaluation System Test (Mini-BESTest)) and gait velocity. Mixed-design analyses of variance were used to determine potential training effects at 6- and 12-month follow-up, and piecewise regression models predicted the rate of deterioration. Results: Seventy-six participants were included at final follow-up. No significant (P > .05) between-group differences remained at either 6 or 12 months following the intervention. The mean Mini-BESTest scores of the training and control group were 19.9 (SD 4.4) and 18.6 (SD 4.3), respectively, at the 12-month follow-up. Gait speed was 1.2 (SD 0.2) m/s in both groups at 12 months. The training group showed a larger deterioration rate per month in balance performance (0.21 point) and gait velocity (0.65 cm/s) than controls (P < .05). Conclusion: These results suggest that training effects diminish within 6 months after balance training, implying that the program may need to be repeated regularly.