ABSTRACT
BACKGROUND: Despite findings from translational and genetic studies in the event-related potential (ERP) literature, the validity and reliability of P50 suppression as a schizophrenia spectrum endophenotype has been questioned. Here, we aimed to examine sensory registration and gating measures derived from P50 and N100 amplitude, as well as N100 area-a novel approach proposed herein-in early psychosis versus health. METHODS: Individuals at clinical high risk for psychosis (CHR; n = 77), first-episode psychosis (FE; n = 52), and healthy controls (HC; n = 65) were assessed in a paired-click auditory ERP paradigm. Eight CHR converted to psychosis (CHRC) and 39 did not (CHR-NC) by 24 months, while 30 CHR were lost to follow-. Group differences, test-retest reliability, and associations with neurocognitive function were assessed in nine ERP measures. RESULTS: Significant differences were observed in N100 S1 amplitude, S1 area, and area difference between HC and FE, as well as in N100 S1 area between HC and CHR, among the total population. Furthermore, significant differences were found in N100 S1 area between HC and CHR-NC (Cliff's delta, Δ = 0.32), as well as in N100 area difference between HC and CHR-C (Δ = 0.55). Both N100 S1 area and area difference demonstrated moderate to acceptable reliability (intraclass correlation coefficients: 0.61-0.78). Processing speed negatively correlated with both N100 S1 area and area difference, while executive function negatively correlated with N100 S1 area alone in CHR and FE. CONCLUSION: Among the ERP measures studied, N100 area measures may serve as a reliable biomarker of aberrant sensory processing and neurocognition in early psychosis.
