S161. DYNAMIC FUNCTIONAL NETWORK CONNECTIVITY COMPARING AUDITORY VERBAL HALLUCINATIONS IN PSYCHOTIC AND NON-PSYCHOTIC SUBJECTS

• Published in: Schizophrenia bulletin Vol. 46; no. Supplement_1; pp. S97 - S98
• Main Authors: Marschall, Theresa, Curcic-Blake, Branislava, Brederoo, Sanne, Sommer, Iris
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 18.05.2020
• Subjects: Hallucinations; Neurobiology; Poster Session I; Psychosis
• ISSN: 0586-7614; 1745-1701; 1745-1707; 1745-1701
• DOI: 10.1093/schbul/sbaa031.227

BackgroundAuditory verbal hallucinations (AVH) are often seen as a hallmark of schizophrenia, but can also occur in the general healthy population. While AVH in non-clinical populations might offer an opportunity to study them in isolation, it remains debatable whether the mechanisms underlying AVH are the same in clinical and non-clinical populations. For example, non-clinical populations are reported to attribute lower emotional valence to their AVH. Such differences in phenomenology are hypothesized to arise from differences on the neurobiological level. With the current study, we employ a data-driven approach to define brain networks involved in AVH in clinical and non-clinical subjects, and test whether dynamic differences in network connectivity exist between these groups.MethodsFunctional magnetic resonance imaging data of 21 non-psychotic individuals and 21 matched psychotic patients with frequent AVH were obtained. During scanning, subjects manually indicated the on- and offset of their AVH. Using independent component (IC) analysis, the data were split into 72 statistically independent spatial maps and their time courses. These time courses were regressed with the AVH time courses. With a one sample t-test on the beta weights, we selected those ICs that related to AVH in both groups for further dynamic functional network connectivity analysis. To identify functional connectivity states, k-means clustering was implemented on correlation matrices acquired using sliding windows. Group differences between these states were determined with two-sample t-tests.ResultsBoth groups experienced AVH during scanning, with a mean number of 24.71 AVH episodes in the clinical and 17.14 episodes in the non-clinical group. We identified seven ICs with time courses significantly related to the occurrence of AVH in both groups. The auditory, sensorimotor, and posterior salience network were positively related to AVH occurrence. The ventral default mode network (DMN), anterior salience network and a network consisting of (para-)hippocampal areas were negatively related to AVH.While in general, networks related to AVH were similar in both groups, a significant difference between the two groups was found in the mean dwell time in states characterized by varying connectivity between these networks. Psychotic patients spent more time in a state of low connectivity (r < 0.055) between all AVH-related networks. Non-psychotic patients dwelled longer in a different state, where some weak correlations between networks were present (.15 > r ≥ .10). Specifically, networks positively related to AVH showed small negative correlations with each other, and a small negative relationship with the DMN. At the same time, the anterior salience network displayed a small positive relationship with the sensorimotor, auditory and posterior salience networks.DiscussionOur findings suggest that similar brain networks underlie AVH in non-psychotic and psychotic individuals, but that the groups differ in terms of connectivity between those networks. Among the involved networks are those typically associated with AVH in psychotic patients, such as the DMN and auditory network. During the experience of AVH, psychotic individuals are more likely to show a state defined by segregation of the AVH-related networks. On the contrary, during AVH non-psychotic individuals are in a state defined by more connectivity between the networks. This suggests that a distinction between clinical and non-clinical AVH may have its neurobiological basis in the extend of disruption of involved network connectivity.