Aberrant brain activation and coupling in Depression – Links between Psychopathology and Neurophysiology

Abstract

Major Depressive Disorder (MDD) is the most common mental disorder and ranging under the top three of the most disabling diseases worldwide. Although various treatments exist for MDD, about 30 to 40 % of the patients don’t respond. A better understanding of the neurobiological correlates of MDD might lead to the development of new and the improvement of existing treatments. The dissertation at hand is dedicated to the aim of a better understanding of aberrant brain functioning and coupling in MDD. Further, we sought to investigate the behavioral and cognitive-affective underpinnings that lead to aberrant brain functioning and coupling in MDD, in terms of depressive rumination. In total this work comprises four studies. In our first study, we investigated the functional connectivity (FC) during resting state (rsFC) and task performance of the Trail Making Test (TMT) in subjects with late-life depression (LLD) and healthy controls (HC). FC was assessed via functional near-infrared spectroscopy (fNIRS) in areas of the cognitive control network (CCN). While we observed an expected pattern of change in FC in the healthy controls with relatively low FC during resting-state and an increase during task-performance, subjects with LLD showed an opposite pattern, with relatively high FC during resting-state and decreases during task-performance. Further, depressed and non-depressed subjects differed significantly during resting-state (LLD>HC) and the executive demanding condition of the TMT (HC>LLD). While these results were interesting from a standpoint of pathophysiological changes in FC, we couldn’t give a final explanation for the observed FC patterns in LLD. As a possible explanation, we assumed that some kind of depressive cognitive process could lead to hyper-connectivity within the CCN during resting-state that further disturbs cortical coupling during task performance. As depressive rumination is such a cognitive process that is common in depression, we developed a resting-state questionnaire to assess state rumination for subsequent studies. In study two, we investigated rsFC within subjects with MDD and HC with a parietal probeset covering parts of the default mode network (DMN), CCN and dorsal attention network (DAN). Further, we investigated in how far state- and trait rumination explained the differences between depressed and non-depressed subjects in rsFC. In contrast to our first study, we observed an opposite pattern of FC differences between the groups: within the parietal cortex, depressed subjects showed reduced FC in comparison to HC in a widespread bilateral network. While state rumination showed rather restricted effects, the measures of trait rumination showed wide-spread effects. Further, FC was negatively correlated with state- and trait rumination. Since our results so far were restricted to non-experimental between-subject associations, that don’t allow the investigation of causal relationships, we further designed a study in which we sought to induce rumination with the Trier Social Stress Test (TSST). In study three, we investigated the hemodynamic changes during the TSST in high and low trait ruminators in the CCN, further, we examined in how far state rumination would be induced through the TSST. Relationships between hemodynamic responses and state rumination were examined with a mediation analysis. As expected, we found increases in state rumination through the TSST. Further, these increases were higher in the high-trait ruminators. On a cortical level, low ruminators showed higher cortical activation in the stress condition than the high ruminators in the right inferior frontal gyrus (IFG). Further, group differences in post-stress state rumination were mediated by the cortical reactivity in this region. Subject with high IFG reactivity showed less state-rumination following the TSST. In study four, we further investigated in the same study cohort, if rsFC before and after stress would show an expected pattern with higher baseline FC in the high trait ruminators and a higher reactivity in rsFC in subjects with high increases in state rumination. As expected, baseline levels of rsFC were increased in the high-ruminators like in our first study for the LDD group. However, although state rumination increased in the high trait ruminators more strongly than in the low trait ruminators, rsFC only increased in the latter group. Since we didn’t observe a co-variation of change scores between rsFC and state rumination, we concluded that the effect of rumination on FC changes would be an indirect one. In the general discussion of this dissertation, I propose a model of indirect prolonged stress effects in high ruminating subjects that lead to higher stress levels and subsequently to permanent changes in FC. This model would explain the observed effects in our study and is in line with current research of FC alterations in chronic stress. I further outline, in how far the presented results and the research of biological underpinnings could improve the current theory development of mental diseases as well as treatment planning.