Objective: To determine whether patients with schizophrenia have abnormalities in the P300 and P50 waves and to quantify the magnitude of any differences from controls.
Method: We conducted a systematic search for articles published between January 1994 and August 2003 that reported P50 or P300 measures in schizophrenic patients and controls. Metaregression analyses were performed using a random effects model. The pooled standardised effect size (PSES) was calculated as the difference between the means of the two groups divided by the common standard deviation.
Results: We identified 46 studies suitable for analysis of P300 measures, including 1443 patients and 1251 controls. There were 20 P50 studies including 421 patients and 401 controls. The PSES for the P300 amplitude was 0.85 (95% CI: 0.65 to 1.05; p<0.001), and for the P300 latency was -0.57 (95% CI: -0.75 to -0.38; p<0.001). The PSES of the P50 ratio was -1.56 (95% CI: -2.05 to -1.06; p<0.001). There were no significant differences between patients and controls in P50 latency. Across-study variations in filters, task difficulty, antipsychotic medication and duration of illness did not influence the PSES significantly.
Conclusions: This meta-analysis confirms the existence of ERP deficits in schizophrenia. The magnitude of these deficits is similar to the most robust findings reported in neuroimaging and neuropsychology in schizophrenia.