BACKGROUND: Increasing evidence for contribution of hemostasis components in multiple sclerosis (MS) has been reported. Hemostasis protein inhibitors display key regulatory roles, extending to regulation of innate immune response and inflammation, and promotion of blood-brain barrier integrity. Whereas the effects on hemostasis of exercise and rehabilitation strategies have been extensively investigated, relationships between MS rehabilitation strategies and hemostasis have not been previously reported. OBJECTIVES: To investigate in MS patients the association between outcomes of rehabilitative exercise and plasma levels of selected hemostasis inhibitors. METHODS: Sixty-one severely disabled progressive-MS (P-MS) patients were randomized in the RAGTIME trial to receive 12 walking session of robot-assisted gait training (RAGT) or conventional overground therapy (CT). Outcome parameters were: timed 25-foot walk test (T25FWT) speed, 6-minute walking test (6MWT), Berg Balance Scale (BBS), and MS impact scale-29 (MSIS-29). Plasma levels of coagulation inhibitors protein S (PS), soluble thrombomodulin (sTM), and tissue factor pathway inhibitor (TFPI) were assayed by multiplex assay and ELISA at 4-time points: baseline (T0), intermediate (T1), end of rehabilitation (T2), 3-month follow-up (T3). Descriptive analysis, trend analysis, Spearman's rank and Pearson's correlations, and multiple regression models were used. RESULTS: Rehabilitative exercises moderately modified plasma protein concentrations. A significant trend to increase was observed for PS (p=0.015) and TFPI (p=0.047) in the whole population, and for PS (p=0.011) in the CT group. Correlation between TFPI and sTM levels was detectable at all time points in the whole P-MS patients and in RAGT group. The correlation between TFPI and PS, present at T0, was lost during the rehabilitation, and recovered at T3 in the whole population and CT group. During rehabilitation, positive variations of TFPI were inversely related with changes in 6MWT in the whole population (r=-0.309, p=0.021), and in the RAGT group (r=-0.51, p=0.004). In all P-MS, PS T0 levels were associated (r=0.379, p=0.004) with increased gait speed, which in the RAGT group was associated both with PS T0 (r=0.378, p=0.040), and sTM T0 (r=0.453, p=0.012). Accordingly, in the regression model including age, sex and EDSS and the stepwise enter of PS T0, higher PS T0 levels predicted increased gait speed in all P-MS (F=3.4, p=0.016) The regression model in the RAGT group indicated that higher PS and sTM T0 levels were both predictors of increased gait speed (F=5.7, p=0.001). CONCLUSIONS: Plasma levels of coagulation inhibitors were related to variations of outcome measurements after high-intensity walking rehabilitation programs. Patients with decreased TFPI levels from T0 to T2 displayed the most significant functional recovery following rehabilitation, and particularly after RAGT. Higher baseline total PS levels were associated with favorable outcomes of rehabilitation therapies in MS. These novel findings, which suggest that plasma levels of hemostasis inhibitors might have implication for rehabilitative therapy options in MS, warrant further investigation.