Neural correlates of theory of mind impairment in multiple sclerosis

Abstract

Despite the gathered knowledge about cognitive impairment in multiple sclerosis (MS), little is known concerning the disease impact on social cognition. An essential aspect for social cognition is theory of mind (ToM), defined as the ability to infer other persons’ mental states. In the present study, we aimed to explore how ToM is affected in MS, particularly if ToM deficits are independent of the classic cognitive impairment associated with the disease; to examine the relationship of ToM and executive functions (EF) performance; and to identify the underlying neural correlates. We enrolled consecutively 60 patients with MS and 60 healthy controls (HC) matched for age, gender, and education. All participants underwent ToM testing (Eyes Test and Videos Test); global cognitive assessment with a standard neuropsychological battery for MS (MACFIMS) in order to classify patients as either cognitively impaired or cognitively intact; tests tapping different processes of executive functions; and 3Tesla brain MRI. Using FreeSurfer software, cortical and subcortical grey matter (GM) volumes were calculated. Tract-based spatial statistics (TBSS) were applied for whole-brain voxel-wise analysis of fractional anisotropy (FA) and mean diffusivity (MD) on normal-appearing white matter (NAWM). We found that patients with MS performed worse on both tasks of ToM compared to HC, i.e. Eyes Test (58.7±13.8% vs. 81.9±10.4%, p<0.001) and Videos Test (75.3±9.3% vs. 88.1±7.1%, p<0.001). There were no significant differences on Eyes Test and Videos Test performance between MS patients with cognitive impairment (n=34; 56.7%) and those with normal cognitive performance (n=26; 43.3%). Moreover, the group of patients without cognitive impairment presented significantly lower scores on both tasks compared to HC. A hierarchical cluster analysis showed that ToM measures were clustered separately from the EF measures, distinguishing three executive clusters (attention/working memory; inhibitory control/shifting ability; verbal initiative/abstract reasoning) and one ToM cluster. ToM performance in MS was positively correlated with the volume of subcortical structures (amygdala, putamen) and cortical regions (entorhinal cortex, fusiform gyrus, superior temporal gyrus, superior parietal gyrus, supramarginal gyrus, medial orbitofrontal cortex, anterior and posterior cingulate gyrus). In regression analysis, amygdala volume was the single predictor of ToM performance. Regarding the TBSS analysis, ToM tests were positively correlated with FA and inversely correlated with MD across widespread NAWM tracts of both hemispheres. The largest effects were for the Eyes Test: body and genu of corpus callosum, fornix, tapetum, uncinate fasciculus, and left inferior cerebellar peduncle; for the Videos Test: genu and splenium of corpus callosum, fornix, uncinate fasciculus, left tapetum, and right superior fronto-occipital fasciculus. Taken together, these results suggest that social cognition is impaired in patients with MS independently of the classic MS-related cognitive impairment and of the EF performance. The social brain network in MS is affected by two different mechanisms: direct damage of the main cortical and subcortical GM nodes, particularly amygdala, and by disconnection between the mentioned nodes caused by injury of the interconnecting white matter tracts.