Visual system assessment for predicting a transition to psychosis

Abstract

The field of psychiatry is far from perfect in predicting which individuals will transition to a psychotic disorder. Here, we argue that visual system assessment can help in this regard. Such assessments have generated medium-to-large group differences with individuals prior to or near the first psychotic episode or have shown little influence of illness duration in larger samples of more chronic patients. For example, self-reported visual perceptual distortions-so-called visual basic symptoms-occur in up to 2/3rds of those with non-affective psychosis and have already longitudinally predicted an impending onset of schizophrenia. Possibly predictive psychophysical markers include enhanced contrast sensitivity, prolonged backward masking, muted collinear facilitation, reduced stereoscopic depth perception, impaired contour and shape integration, and spatially restricted exploratory eye movements. Promising brain-based markers include visual thalamo-cortical hyperconnectivity, decreased occipital gamma band power during visual detection (MEG), and reduced visually evoked occipital P1 amplitudes (EEG). Potentially predictive retinal markers include diminished cone a- and b-wave amplitudes and an attenuated photopic flicker response during electroretinography. The foregoing assessments are often well-described mechanistically, implying that their findings could readily shed light on the underlying pathophysiological changes that precede or accompany a transition to psychosis. The retinal and psychophysical assessments in particular are inexpensive, well-tolerated, easy to administer, and brief, with few inclusion/exclusion criteria. Therefore, across all major levels of analysis-from phenomenology to behavior to brain and retinal functioning-visual system assessment could complement and improve upon existing methods for predicting which individuals go on to develop a psychotic disorder.

Fig. 1. Visual tasks that may elicit behavioral markers of early psychosis.

A Never-medicated first-episode psychosis patients (FEPs) have superior contrast sensitivity at lower spatial frequencies compared to healthy controls (HCs) [32]. B Never-medicated FEPs have a diminished ability to identify the circle that floats in stereo [35]. C FEPs need more of a temporal interval (SOA) between the target and a subsequent grid-like mask to achieve 75% accuracy on a Vernier discrimination task [42]. D Schizophrenia patients of varying illness durations benefit less from high-contrast collinear flankers when attempting to detect a central low-contrast target relative to healthy and clinical controls [45, 47]. E For FEPs, discriminating fat and thin illusory shapes is harder than discriminating left and right rotated pac-men; for HCs, the opposite is true [48]. F Compared to healthy controls, FEPs can tolerate fewer background “noise” elements when attempting to detect a circular chain of elements [50]. G Independently of illness duration, SZ patients exhibit spatially restricted eye movement patterns when freely viewing complex images [51].

Fig. 2. Visually oriented brain-based markers of early psychosis.

A Visual thalamo-cortical hyperconnectivity occurs in schizophrenia patients relative to HCs (Figure panel shows the thalamic seed; adapted with permission from [58]). B Similar hyperconnectivity patterns have been observed in CHR patients who convert versus those who do not convert to a psychotic disorder (Adapated with permission from [61]). C Different investigators have found comparable patterns (after FDR correction) in early-stage non-affective psychosis patients relative to HCs, with one-third of the visual nodes being hyperconnected to the thalamic node (Adapted with permission from [60]). D Briefly flashed checkered images have elicited a muted P1 amplitude in FEPs relative to HCs (Adapted with permission from [65]). E, F During a change-in-motion detection task, gamma band activity within ten occipital regions was smallest in FEP subjects, intermediate in CHR patients, and greatest in psychiatric patients not meeting CHR criteria (non-CHR) (Adapted with permission from [68]).

Fig. 3. Electroretinographic markers of early psychosis.

A Two example waveforms that have distinguished schizophrenia patients from healthy controls and mood disorder patients (photopic flicker response not shown). B Using a portable handheld device, light-adapted (photopic) retinal functioning of different cell types can be measured within 15 min. C Four ERG variables that have significantly distinguished schizophrenia patients from well-matched healthy controls and people with major depression (the most common final diagnosis among those at clinical high risk for psychosis; n = 25/group). The a- and b-wave amplitudes were measured under three photopic viewing conditions (P1, P2, P3); the photopic flicker response was measured under a fourth condition (for details see [81]). All effects were in a direction that could be anticipated from past research. *p < 0.05, **p < 0.01, ***p < 0.001.