Sleep deprivation (SD) is often associated with significant shifts in mood state relative to baseline functioning. Prior work suggests that there are consistent trait-like differences among individuals in the degree to which their mood and performances are affected by sleep loss. The goal of this study was to determine the extent to which trait-like individual differences in vulnerability/resistance to mood degradation during a night of SD are dependent upon region-specific white and grey matter (WM/GM) characteristics of a triple-network model, including the default-mode network (DMN), control-execution network (CEN) and salience network (SN). Diffusion-weighted and anatomical brain data were collected from 45 healthy individuals several days prior to a 28-h overnight SD protocol. During SD, a visual analog mood scale was administered every hour from 19:15 (time point1; TP1) to 11:15 (TP17) the following morning to measure two positive and six negative mood states. Four core regions within the DMN, five within the CEN, and seven within the SN were used as regions of interest (ROIs). An index of mood resistance (IMR) was defined as the averaged differences between positive and negative mood states over 12 TPs (TP5 to TP16) relative to baseline (TP1 to TP4). For each ROI, characteristics of WM – quantitative anisotropy (QA) and mean curvature index (WM-MCI), and GM – cortical volume (CV) and GM-MCI were estimated, and used to predict IMR. WM characteristics, particularly QA, of all of regions within the DMN, and most of the regions within the CEN and SN predicted IMR during SD. In contrast, most ROIs did not show significant association between IMR and any of the GM characteristics (CV and MCI) or WM MCI. Our findings suggest that greater resilience to mood degradation induced by total SD appears to be associated with more compact axonal pathways within the DMN, CEN and SN.
Bajaj, S., & Killgore, W. D. S. (2019). Vulnerability to mood degradation during sleep deprivation is influenced by white-matter compactness of the triple-network model. NeuroImage, 202. https://doi.org/10.1016/j.neuroimage.2019.116123