Is BDNF related to spatial-temporal gait parameters in people with multiple sclerosis? An observational study.

Background: It has been suggested that the protein Brain-derived Neurotrophic Factor (BDNF) plays a neuroprotective role in people with multiple sclerosis (pwMS). Also, BDNF seems to play a role in cognition performance. In the same line, gait in pwMS requires a higher cognitive resource, mainly during complex walking. Thus, maybe BDNF could be related to gait in pwMS.

Objective: To investigate the relationship between BDNF and gait spatial-temporal parameters during unobstructed and obstructed conditions and the Timed Up and Go (TUG) in pwMS and healthy controls (HC).

Methods: The study included 20 pwMS (11F/9M, 33.1±7.5 years, Expanded Disability Status Scale- EDSS 2.2±1.2) and 18 HC (13F/5M, 35.5±5.9 years). Both groups performed 20 gait attempts in two conditions: unobstructed walking (10 trials) and avoiding an obstacle. The obstacle was 15 cm in height and made of foam material. The BDNF serum concentration was collected with participants in fasting and completed before the clinical, gait, and mobility assessments. Clinical variables included the Symbol Digit Modality Test (SDMT), the Fatigue Severity Scale (FSS), and the International Physical Activity Questionnaire (IPAQ- short version). Associations between BDNF and spatial-temporal gait parameters, clinical variables, and TUG were determined by Pearson/Spearman correlations with Bonferroni's correction being applied (p<0.0013). Gait was compared by a two-way, repeated-measures ANOVA (group and condition) to characterize our cohort.

Results: Reduced BDNF was observed for pwMS (41.66±4.45 ng/ml) in comparison with HC (61.67±7.07, p<0.001). However, although some correlations presented a moderate correlation between BDNF with gait variables, the correlations didn't reach a significant p-value after Bonferroni's correction. Lastly, pwMS presented shorter step length and slower step velocity for both gait conditions, with more evidence for obstacle conditions. Only pwMS changed gait behavior from unobstructed walking to obstacle avoidance conditions (i.e., reduced step length and velocity and increased step duration).

Conclusion: BDNF is not related to either clinical (i.e., EDSS, SDMT, FSS, or IPAQ) or gait parameters in pwMS and HC, even in a condition involving higher cognitive demand. These results may suggest that BDNF does not play a role in these parameters' performance.