Cognitive training improves working memory, processing speed, and neural efficiency in multiple sclerosis

Introduction: Multiple sclerosis (MS) is a demyelinating disease of the central nervous system (CNS) that features multifocal white and gray matter lesions. Cognitive impairments in working memory (WM) and processing speed (PS) are common features of MS. Cognitive training has shown promise for improving cognitive deficits in MS. The purpose of the present study was to investigate the effects of WM and PS training on performance of a 2-back WM task and 4 × 4 matrix Visual Search (VS) task and to determine the effectiveness of training as measured by a number of purported behavioral indices of neural efficiency, such as RT variability. Method: Forty three Relapsing-Remitting MS participants (PwMS) completed the study. Participants visited the laboratory for pre-testing, were assigned to a training group (WM group,VS group), or a MS control group (no-training) and returned to the laboratory for post-testing after approximately 4–5 weeks of home training. At pre- and post-testing, all participants were administered the WM and VS Tasks. Behavioral measures of accuracy, reaction time (RT), and RT-variability were obtained to these tasks. Results: Analyses revealed that the WM training group had significantly improved behavioral measures at posttest on the 2-back WM task compared to the other two groups, and improvement on the VS task, indicative of transfer of training from WM to VS. The VS training group showed significantly improved behavioral measures on the VS task at posttest, and improved RT on the 2-back task at posttest. The control group improved significantly only for accuracy on the VS Task from pre- to posttest and failed to show significant improvements on any RT measures for either the VS or WM tasks. RT variability (purported markers of neural efficiency) provided unique information related to training effects and signified improved neural efficiency due to training. Conclusion: The findings suggest that training, particularly on WM, leads to increased neural efficiency, reflected by decreased RT variability, increased PS, and transfer of training across tasks.