Additional VR-based training improves upper extremity functions in subacute stroke patients: a single-blinded pairwise-matched trial

Background Stroke is the leading cause of disability worldwide. Upper extremity paresis is the most common functional consequence, affecting more than half of all stroke survivors. Research has shown that an adequate therapy should begin in the sub-acute stage, but also that an enhanced intensity and frequency of therapy can positively affect patients’ recovery curve. Therefore, here we assessed whether an additional VR-based training can be beneficial for recovery of stroke patients, with particular emphasis on upper extremity functions. Methods The study was organized as prospective and single-blinded (analysis). Two groups of pairwise-matched subacute stroke patients with arm paresis were recruited at our rehabilitation clinic while controlling for gender, age, sidedness and modified Rankin scale (mRS). Both groups – conventional therapy (CT) and conventional therapy plus virtual reality (CT + VR) – received 30–45 min of conventional therapy on 3 to 4 days/week over 4 weeks; in addition to that, the CT + VR group received 3 times per week a specially designed VR-based training for upper extremity. Data acquisition was performed within 24 h before the baseline and after the training has ended. Main outcomes were patients’ performance on Fugl-Meyer test for upper extremity (FME), Box-and-Block test, hand dynamometry and Functional independence test (FIM). Results Twenty-two subjects aged 57–85 were pairwise-matched and assigned to the conventional therapy (CT) group ( n = 11, 67.82 ± 8.69 years; three females) and the conventional therapy plus virtual reality (CT + VR) group ( n = 11, 70.45 ± 6.79 years; three females). No difference in gender, age, sidedness, mRS and mini-mental status examination (MMSE) existed between the two groups. The CT + VR group showed significantly better improvements over time on FME (44.3 ± 7.8 to 58.7 ± 11.2 vs. 42.1 ± 6.2 to 49.5 ± 10.9; p = 0.009) and FIM (90.1 ± 18.0 to 118.1 ± 6.9 vs. 105.0 ± 12.4 to 110.6 ± 12.6; p < 0.001), compared to the CT group. Other tests revealed no significant differences. Discussion As hypothesized, an additional immersive VR-based training can be beneficial for stroke patients suffering from upper extremity deficiency. Nevertheless, the principle of specificity could be observed, with only trained functions being associated with improvements on FME and FIM. Future studies with larger sample of participants are required to confirm these findings.