The impact of visual dysfunctions in recent-onset psychosis and clinical high-risk state for psychosis

Abstract

Subtle subjective visual dysfunctions (VisDys) are reported by about 50% of patients with schizophrenia and are suggested to predict psychosis states. Deeper insight into VisDys, particularly in early psychosis states, could foster the understanding of basic disease mechanisms mediating susceptibility to psychosis, and thereby inform preventive interventions. We systematically investigated the relationship between VisDys and core clinical measures across three early phase psychiatric conditions. Second, we used a novel multivariate pattern analysis approach to predict VisDys by resting-state functional connectivity within relevant brain systems. VisDys assessed with the Schizophrenia Proneness Instrument (SPI-A), clinical measures, and resting-state fMRI data were examined in recent-onset psychosis (ROP, n = 147), clinical high-risk states of psychosis (CHR, n = 143), recent-onset depression (ROD, n = 151), and healthy controls (HC, n = 280). Our multivariate pattern analysis approach used pairwise functional connectivity within occipital (ON) and frontoparietal (FPN) networks implicated in visual information processing to predict VisDys. VisDys were reported more often in ROP (50.34%), and CHR (55.94%) than in ROD (16.56%), and HC (4.28%). Higher severity of VisDys was associated with less functional remission in both CHR and ROP, and, in CHR specifically, lower quality of life (Qol), higher depressiveness, and more severe impairment of visuospatial constructability. ON functional connectivity predicted presence of VisDys in ROP (balanced accuracy 60.17%, p = 0.0001) and CHR (67.38%, p = 0.029), while in the combined ROP + CHR sample VisDys were predicted by FPN (61.11%, p = 0.006). These large-sample study findings suggest that VisDys are clinically highly relevant not only in ROP but especially in CHR, being closely related to aspects of functional outcome, depressiveness, and Qol. Findings from multivariate pattern analysis support a model of functional integrity within ON and FPN driving the VisDys phenomenon and being implicated in core disease mechanisms of early psychosis states.

Fig. 1. Prominent example of visual distortions illustrated by a patient with a psychotic disorder.

Copyright courtesy to the artist.

Fig. 2. Summary of VisDys item means and underlying component structure by group.

Displays degrees of severity for each of the 14 VisDys items by group derived from SPI-A (a) and visualization of the number of components underlying VisDys sum score for ROP and CHR groups (scree plots from PCA analyses in b, c, respectively). For more details see Tables S2 and S3.

Fig. 3. Most reliable intrinsic brain activity connectivities predicting visual dysfunctions.

Connectivities (lines) between regions of interest (spheres) derived from resting-state brain activity were identified as classifying the occurrence of visual dysfunctions (VisDys⁺ vs. VisDys^–) in patients with recent onset psychosis (ROP) and subjects at clinical high risk (CHR). Depicted are the most reliable connectivities, within the occipital network (ON) in ROP (a) and CHR groups (b). For the combined ROP + CHR group, connectivities within the frontoparietal network (FPN) and the combined ON-FPN are shown in c, d, respectively. For a list of all cross-validation ratios (CVs) see Tables S9–S12.