Background: The aim of this study was to investigate whether P300 could predict the short-term prognosis of subjects at clinical high risk (CHR) for psychosis who do not convert to psychotic disorder (non-converters).

Method: CHR subjects were examined with auditory P300 at baseline, and their clinical state was regularly assessed up to 2 years. 45 CHR non-converters were divided into remitter and non-remitter groups. Repeated-measures analysis of variance (ANOVA) was performed to compare baseline P300 between the two groups. Multiple regression analysis was used to identify factors predicting symptomatic or functional improvement in CHR subjects during the follow-up period.

Results: There were no group differences in P300 amplitude or latency between CHR remitters and non-remitters. In the multiple regression analysis, P300 amplitude at Pz (β=0.206, 95% confidence interval [95CI]=0.035 to 0.567, p=0.028) significantly predicted later amelioration of the Scale of Prodromal Symptoms (SOPS) negative symptoms. Improvement in SOPS general symptoms was significantly predicted by P300 amplitude at Pz (β=0.255, 95CI=0.065 to 0.455, p=0.010) and mood stabilizer use (β=0.199, 95CI=0.081 to 4.154, p=0.042).

Conclusions: These results indicate that P300 may be a possible predictor of improvement in negative and general symptoms in CHR non-converters. Our findings support the recommendation that a broader concept of assessment guidelines is needed to forecast clinical outcome and provide appropriate interventions for CHR non-converters.