A limitation of the above study is that P3 amplitudes were measured to target stimuli in two-stimulus oddball tasks, which is not ideal for measuring the frontocentral P3a. We therefore undertook a study in which brain ERPs in depressed patients and controls were measured during a novelty oddball task (Friedman et al., 1993). This was designed to evaluate whether depressed patients differ from controls in the novelty P3 associated with orienting of attention or the target P3b associated with resource allocation and memory operations. Given evidence for the role of prefrontal and anterior cingulate cortex in depression (Drevets et al., 1997; Bremner et al., 2004; Siegle et al., 2004) and in the generation of P3a or novelty P3 (Knight and Scabiani, 1998; Dien et al., 2003), we predicted that the frontocentral novelty P3 would be reduced in depressed patients when compared to healthy controls, whereas there would be less difference between these groups in the parietal P3b to target stimuli.
