Working memory performance and thalamus microstructure in healthy subjects.

Research on the neural basis of working memory (WM) has generally focused on cortical regions, specifically frontal and parietal areas. Comparatively, evidence of a possible involvement of deep gray matter structures, that are parts of cortico-cortical circuits linking anterior and posterior cortical areas, is far less clear. The goal of the present study is to test the hypothesis that individual structural variations within deep gray matter structures may affect the cortical networks involved in WM. To this aim, a large sample (n=181) of healthy subjects underwent a high-resolution structural magnetic resonance imaging (MRI) and a diffusion tensor imaging (DTI) scan protocol. Data of micro- (mean diffusivity, MD) and macro- (volume) structural variations of six bilateral deep gray matter structures (thalamus, caudate nucleus, putamen, hippocampus, amygdala and pallidum) and lateral ventriculi volume were analyzed in association with score in a WM (the so-called n-back task) and other neuropsychological tasks. Results showed that increased MD of bilateral thalami was the only structural parameter that significantly correlated with reduced WM performance. In particular, a voxel-by-voxel analysis revealed that the greater percentage of voxels showing significant anticorrelation between WM score and MD values were localized in those thalamic nuclei projecting to prefrontal and posterior parietal cortices. Results highlight the specific involvement of thalamus microstructure, not volume, in modulating WM performance, possibly by regulating the connections among cortical areas that are recruited during WM tasks.