Aberrant Effective Connectivity During Eye Gaze Processing Is Linked to Social Functioning and Symptoms in Schizophrenia

Abstract

Background: Patients with schizophrenia show abnormal gaze processing, which is associated with social dysfunction. These abnormalities are related to aberrant connectivity among brain regions that are associated with visual processing, social cognition, and cognitive control. In this study, we investigated 1) how effective connectivity during gaze processing is disrupted in schizophrenia and 2) how this may contribute to social dysfunction and clinical symptoms.

Methods: Thirty-nine patients with schizophrenia/schizoaffective disorder (SZ) and 33 healthy control participants completed an eye gaze processing task during functional magnetic resonance imaging. Participants viewed faces with different gaze angles and performed explicit and implicit gaze processing. Four brain regions-the secondary visual cortex, posterior superior temporal sulcus, inferior parietal lobule, and posterior medial frontal cortex-were identified as nodes for dynamic causal modeling analysis.

Results: Both the SZ and healthy control groups showed similar model structures for general gaze processing. Explicit gaze discrimination led to changes in effective connectivity, including stronger excitatory, bottom-up connections from the secondary visual cortex to the posterior superior temporal sulcus and inferior parietal lobule and inhibitory, top-down connections from the posterior medial frontal cortex to the secondary visual cortex. Group differences in top-down modulation from the posterior medial frontal cortex to the posterior superior temporal sulcus and inferior parietal lobule were noted, such that these inhibitory connections were attenuated in the healthy control group but further strengthened in the SZ group. Connectivity was associated with social dysfunction and symptom severity.

Conclusions: The SZ group showed notably stronger top-down inhibition during explicit gaze discrimination, which was associated with more social dysfunction but less severe symptoms among patients. These findings help pinpoint neural mechanisms of aberrant gaze processing and may serve as future targets for interventions that combine neuromodulation with social cognitive training.

Keywords: DCM; Face processing; Gaze; Schizophrenia; Social cognition; fMRI.

Figure 1.. Face stimuli and design of the gaze processing task

Note. Participants completed an eye-contact detection task in which they determined whether individuals pictured in a series of face stimuli were looking at them or not (i.e., “explicit gaze discrimination”). Color photos of six actors with 9 different gaze angles (A) were used. Angles represent levels of “eye-contact signal strength”: 0° – direct, 3.75°, 7.5°, 11.25°, 15°, 18.75°, 22.5°, 26.25°, 30° – most averted. As a control task that involved implicit gaze processing, participants determined whether faces were male or female. Eyes and Gender trials were presented in alternating blocks with a fixation block between each task block (B). The task was organized into 6 runs, with each run containing 6 task blocks (3 gaze and 3 gender) and 5 fixation blocks. Within each block, the gaze angle and gender of face stimuli were pseudo-randomized. Each face was presented for 1.5s and separated from the next face by a random jitter (1.6 – 3.9 s). There was a total of 216 trials (6 runs x 3 blocks x 2 conditions x 6 trials). After psychophysical curve-fitting based on task performance data, two metrics were obtained for each participant: threshold and slope (C). Face images were taken from George et al. (2001) [86]. The procedure to generate stimuli for various gaze angles is described in Lasagna et al. (2020).

Figure 2.. Brain nodes and our model for dynamic causal modeling (DCM)

Note. A) Based on the GLM result for our Gaze – Gender contrast across all participants, 4 nodes were selected for subsequent DCM analysis: V2 visual cortex (Vis), posterior superior temporal sulcus (pSTS), inferior parietal lobule (IPL), and posterior medial frontal cortex (pMFC). B) The model used for our DCM. Black arrows indicate connections during all face events (Matrix $A$), while green dots indicate the connections that were allowed to be modulated by explicit gaze discrimination (Matrix $B$). The red arrow indicates driving input (Matrix $C$). C) The DCM in terms of on/off parameters specified in Matrices $A$, $B$, and $C$.

Figure 3.. Posterior probability density plots of the psychophysical gaze perception metrics

Note. Dashed vertical lines and numbers at the top indicate median values of Markov Chain Monte Carlo (MCMC) samples. Numbers at the bottom right (bottom panel) indicate the posterior probability of the HC - SZ group difference >0. (This probability is represented by the area under the curve to the right of the solid vertical line)

Figure 4.. Effective connectivity parameters

Note. Matrix A represents connectivity during all face events. Matrix $B$ represents connectivity changes due to explicit gaze discrimination. For group differences, red indicates parameters that are greater in SZ (vs. HC), whereas blue indicates parameters with lower values in SZ. For clarity, only parameters whose values are non-zero with posterior probability >75% are shown.

Figure 5.. Canonical correlation between effectivity connectivity and A) social dysfunction or B) symptoms

Note. Sample for the social dysfunction CCA included both SZ and HC participants, whereas the symptom severity CCA included only SZ. Ovals represent predictors (left) and criterion variables (right). Vis = V2 visual cortex; IPL = inferior parietal lobule; pSTS = posterior superior temporal sulcus; pMFC = posterior medial frontal cortex; MSCEIT Total = Mayer–Salovey–Caruso Emotional Intelligence Test Emotion Intelligence Quotient, RME = Reading the Mind in the Eyes Task accuracy; ER-40 = Penn Emotion Recognition Task accuracy; QCAE = total score on the Questionnaire for Cognitive and Affective Empathy, SSPA = total score on the Social Skills Performance Assessment; UPSA = total communication scale score on the UCSD Performance-Based Skills Assessment; SAPS/SANS = Scale for the Assessment of Positive/Negative Symptoms.

Figure 6.. Scatter plot visualizations of the canonical correlations between effective connectivity and A) social dysfunction or B) symptoms

Note. DCM connections that differentiated the two diagnostic groups were also significantly associated with social dysfunction (A) and clinical symptom severity (B), within the SZ group. **p < .01