SA74. Visual Cortex Gamma-Band Power is Reduced in Individuals at Ultra-High Risk for Schizophrenia: An MEG Study

Abstract

Abstract Background: Emerging evidence suggests that schizophrenia (ScZ) is associated with aberrant neural oscillatory activity, in particular at γ-band frequencies (>30 Hz). Our group has reported several studies showing impaired generation of high-gamma band oscillatory activity (~50–120 Hz) in chronic ScZ patients as well as in unmedicated patients with first-episode psychosis (FEP). Currently, however, it is unclear whether gamma-band activity is already impaired in individuals at Ultra High Risk (UHR) for schizophrenia. The current MEG study investigates this question, using a visual task that is known to induce strong high-γ band activity over visual cortical areas. Methods: MEG data were recorded, while 50 UHR and 30 healthy control (CON) participants performed an inward moving grating task. Participants fixated on sine-wave circular gratings and detected (by button-press) acceleration of inward movement, occurring at an unpredictable moment during trial presentation. MEG analysis focused on frequency domain reconstructed MEG data collected from central nodes placed in 80 (sub)cortical regions defined within the AAL-atlas, using LCMV beamformer source-analysis algorithms to optimally suppress both sensor-level noise sources and contributions from neighboring brain areas. Results: Independent of group (i.e., across all participants), 14 AAL-atlas-derived regions in the visual cortex showed significantly increased sustained gamma-band (53–63 Hz) and significantly decreased sustained alpha-band (8–12 Hz) activity during stimulus presentation, compared to baseline activity (False-Discovery-Rate corrected for multiple comparisons). Subsequently tested group effects (UHRs vs. CONs) across these 14 visual regions and across a time interval between 200 and 1200 ms postgrating onset revealed significantly decreased gamma-band responses for UHR, but no group differences in alpha-band responses. In addition, UHR participants’ performance was characterized by elevated error rates, compared to CON participants. Conclusion: Gamma-band visual-cortex responses during perception are already impaired in individuals at UHR for psychosis development and therefore represent a potential candidate biomarker for psychosis prediction.