Effects of high-intensity robot-assisted training in hand function recovery and adl independence in individuals with multiple sclerosis: A randomized controlled single-blinded trial

Introduction: Functional impairment of UL plays an important role in the daily activities in patients with Multiple Sclerosis (MS) and strongly influences their quality of life (QoL) [1]. Despite this, limited research is dedicated specifically to UL performance and training in persons with MS [2]. Although the integration of robotics into rehabilitation is a promising development, till now there is a lack of evidence on the effects of robot-aided therapy on arm motor performance, functional capacity and disability in patients with MS. The aims of the study focused on patients with MS are: to compare the effectiveness of high-intensity robot-assisted training vs. conventional treatment on sensorimotor hand recovery, ADL and QoL; to explore the neuromuscular pattern of UL recovery by using sEMG. Methods: This single-blind, randomized, controlled trial involved 43 outpatients (age mean 50 ± 11 yrs, EDSS 2–8) with MS (relapsing-remitting). Patients were randomized in two groups: experimental group (n = 21; Robot-assisted training) and control group (n = 22; conventional treatment). Both treatments consist in 10 sessions (45′, twice per week, 5 weeks). All patients were evaluated in single blind pre-treatment (T0), post-treatment (T1) and 1-month follow-up (T2). Primary outcome: Fugl-Meyer Assessment Motor Scale (FMA). Secondary outcome: Action Research Arm test (ARAT), Tremor Severity Scale, Nine Hole Peg Test (NHPT), Motricity Index (MI), Visual analog Scale for tiredness and fatigue and sEMG. sEMG was performed on 7 UL muscles of the paretic side (deltoid anterior/posterior, biceps/triceps brachii, grand pectoral, extensor/flexor radialis carpi) during two tasks: “Hand to mouth” and “Grasp a 10cm-wooden-block” (ARAT sub-test). 14 healthy age matched controls underwent one session of sEMG acquisition to collect normative data. Non-parametric statistical tests were used (SPSS, Ver. 23.0). sEMG signals were computed by means of custom software routines implemented under the Matlab environment (Matworks Inc., Natick, MA, USA). The sEMG signals were processed by using an adaptive pre-whitening filter and the approximated generalized likelihood-ratio (AGLR) algorithm in order to detect the muscle activity. The onset and offset of muscle activity were also analyzed as percentage of the full movement (Fig. 1). Results: Preliminary analysis on 34 patients (17 CG and 17 EG) showed no significant difference between groups on primary and secondary outcomes over time. Within group comparisons showed significant effect between T0 and T2 in EG on FMA (p < 0.05), ARAT (p < 0.05), NHPT (p < 0.05) and MI (p < 0.05). Preliminary results suggest that proximal muscle impairment (deltoid anterior/posterior, biceps brachii and grand pectoral) may account for UL impairment. Discussion: A robot-assisted training for the recovery of hand dexterity can promote functional recovery of UL in patients affected by MS. The analysis of sEMG recorded in patients with MS may contributes to the understanding of the mechanism underlying UL recovery, which is crucial to identify effective strategies for UL rehabilitation.