No evidence for an effect of a working memory training program on white matter microstructure

Numerous studies demonstrate that working memory training benefits cognitive performance; however, such findings are controversial and the field is polarized. Despite a multitude of behavioural studies, investigation into the neurobiological mechanisms underlying working memory training remains limited. Diffusion tensor imaging (DTI) allows for the visualization of white matter structure and is a promising approach to identify neurobiological change associated with working memory training. In this study, 25 healthy community dwelling adults (age 18–40 years) underwent neuroimaging before and after completing a 6-week home- and web-based complex working memory training program. An active control group (n = 24) completed 6-weeks of processing speed training using the same computerized platform and imaging protocol. Voxel-wise statistical analysis of the DTI data was conducted with FSL's Tract-Based Spatial Statistics (TBSS) to assess change in several metrics over the course of the intervention. Additionally, averaged whole-brain values (i.e., throughout all cerebral white matter) for each DTI metric were examined via one-way analysis of covariance (ANCOVA). The participants' scores on the training tasks significantly improved post-training. The training was not associated with improved scores on non-trained cognitive measures. The working memory training also did not alter regional or global white matter microstructure, as measured by fractional anisotrophy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD). Consistent with our previous published findings on brain activity and grey matter structure in the same study sample, the present study does not suggest that home- and web-based complex working memory training alters white matter microstructure.