Association between retinal and cortical impairments in schizophrenia with visual hallucinations : an electrophysiological study of the visual processing

Abstract Electrophysiological deficits in the visual cortical processing are reported in schizophrenia. Specifically, studies support the hypothesis of a magnocellular impairment in the psychiatric illness. However, recent findings reported electrophysiological anomalies as early as retina. Hence, question arises about the link between these retinal and cortical alterations, especially during magnocellular biased conditions among patients with schizophrenia. Their association with visual symptoms such as visual hallucinations was also investigated in this population. We recorded the P100 amplitude and latency in EEG during the projection of low or high spatial frequency gratings (LSF or HSF ; 0.5 or 15 cycles/degree) presented statically or dynamically (Temporal Frequency TF : 0Hz or 8Hz). We recruited 29 healthy controls (HC, n = 29) and 21 patients with schizophrenia (SZ, n = 21) divided in two subgroups according to the presence or absence of history of visual hallucinations : VH group (n = 9) and auditory hallucinations or no hallucinations group (AHNH group, n = 12). We compared P100 results to former results regarding retinal ganglion cells activity (N95) and visual cognition performances in participants 1. Data were compared between groups by repeated measures ANCOVA, linear regression analyses and mediation analyses. First, analysis showed a decreased P100 amplitude and an increased P100 latency in SZ compared to HC (p < 0.05). Second, analysis reported main effects of SF and TF without group interaction. More, P100 latency was correlated with the VOSP object score and the N95 latency in SZ group. Third, VH group reported an increased P100 latency compared to HC (p < 0.05). P100 latency was also correlated with the VOSP object score and the N95 latency in the VH group. Finally, we found a partial mediation between the P100 latency, N95 latency and VOSP object score in the VH group. P100 alterations in SZ are consistent with the early visual cortical processing deficit shown in literature. Importantly, these deficits seem to be not related with the magnocellular processing but appears to be associated with previous retinal measurements. Thereby, impairments in N95 and P100 were found only in VH group and were associated with the VOSP object score, thus supporting the role of retina in VH. Studies with coupled ERG-EEG measures are now required to clarify these findings.