Background: Upper extremity hemiparesis is one of the most common post-stroke disabilities requiring rehabilitation. Objective: To compare the effects of virtual and real boxing training in addition to neurodevelopmental treatment on the upper extremity, balance, and cognitive functions in hemiparetic stroke patients. Methods: Forty hemiparetic stroke patients were assigned to either real boxing group-RBG (n=20) or virtual boxing group-VBG (n=20), for a total of 24 sessions (3 sessions/week for 8 weeks). The primary outcome was upper extremity motor ability (Wolf Motor Function Test-WMFT). The secondary outcomes were arm-hand dexterity (Manual Dexterity Test-MMDT), goal-oriented performance (Video Boxing Analysis-VBA), balance functions (Fullerton Advanced Balance Scale-FAB-T), and cognitive functions (Addenbrooke's Cognitive Examination-Revised-ACE-R). Results: There was small treatment effect on ACE-R, small-medium effect for WFMT and MMDT and large effect on bilateral punching time [VBA (Cohen's d- VBG=0.83; RBG=0.95)] and balance [FAB-T (Cohen's d - VBG=0.89; RBG=0.82)] after treatment in both groups. No significant differences were found for training effects between the groups for upper extremity functions [WMFT (p=0.799; Cohen's d=-0.07), MMDT-PT (p=0.327; Cohen's d=-0.10), MMDT-THTPT (p=0.779; Cohen's d=-0.17) and VBA bilateral punch number (p=0.068; Cohen's d=0.15)], balance functions [FAB-T (p=0.602; Cohen's d=-0.19)] and cognitive functions [ACE-R total (p=0.947, Cohen's d=0.09)]. Conclusion: The study showed that virtual and real boxing training methods, in addition to neurodevelopmental treatment, are effective in improving upper extremity, balance, and cognitive functions in patients with hemiparetic stroke. The training effects were higher on bilateral punching time and balance functions for both groups. There was no superiority of either approach.
Keywords: Stroke; boxing; cognitive functions; postural balance; upper extremity functions; virtual reality.