Relations of electrophysiological measures of visual function and neuropsychological function in schizophrenia

Abstract

Background: Visual dysfunction is established in schizophrenia. This study investigated relations of electrophysiological measures of visual function and neuropsychological measures to evaluate visual processing deficits in schizophrenia and underlying excitatory/inhibitory processes. Methods: Participants were 40 patients with schizophrenia and schizoaffective disorder and 40 age-matched healthy controls. Steady-state visual evoked potentials ( ss VEPs) were elicited by luminance-modulated bright and dark isolated-checks (icVEPs) and provided measures of amplitude, noise, signal-to-noise ratio (SNR), and phase. A biophysical model applied to contrast-response functions of ss VEPs to dark ic VEPs provided estimates of contrast gain, gain control, and time constants. Lateral processes were examined by use of ssVEPs to contrast-reversing radial (partial-windmill [PW], windmill-dartboard [WD]) stimuli. Nonlinear cortical intermodulation responses were studied using two-sinusoid contrast-reversal of a dartboard pattern. Linear mixed-effects modeling examined group differences in ic VEPs. Predictive accuracy for group classification was estimated using receiver-operating-characteristic (ROC) curve analyses. Relationships between neurocognition (MATRICS/WAIS-III), ssVEP responses, and biophysical model parameters were explored. Results: Patients showed significantly reduced SNRs due to reduced amplitudes and similar noise levels to controls. Patients showed significantly reduced contrast gain, similar contrast gain control to controls, and longer time constants for low contrast. Patients exhibited significantly reduced coherence of the 2nd harmonic responses in PW . Patients exhibited significantly reduced SNR of the 2 Hz difference intermodulation component for the high and low stimulus frequency pairs. Amplitude response to bright checks at 32% DOM produced maximal group separation. Neuropsychological measures of processing speed were significantly correlated with SNR of the two-sinusoid stimuli but not with time constants. Conclusions: Results support impaired contrast and temporal processing and decreased cortical excitation in schizophrenia and provide evidence for the VEP technique and the biophysical model parameters as a diagnostic tool. Further identification of parameter characteristics related to underlying perceptual deficits may inform treatments.