Biol Psychol. 2009 Jul; 81(3): 169-76Vanderhasselt MA, De Raedt RRecently, cognitive control deficits have been explored as a possible underlying trait abnormality in depression. In this study, cognitive control was investigated in homogeneous samples of never depressed controls, formerly depressed patients who had experienced a maximum of two depressive episodes and formerly depressed patients who had experienced at least three depressive episodes. We measured event related potentials (ERP) during a modified Stroop task in a sample of twenty-five formerly depressed patients and thirteen controls. Using this Stroop task, we manipulated subjects' cognitive control for congruent and incongruent trials. These Stroop manipulations have generally been related to two post-stimulus ERP components related to cognitive control, the N2 and the N450. Behavioural measurements of reaction times and error rates did not indicate a difference in cognitive control between the samples. Left frontal N450 amplitude for the high conflict (HC) condition differentiated the three groups. While a N450 conflict-related modulation was significant in the healthy comparison group, such an effect was significantly reduced in the remitted depressed patients, particularly in the high recurrent group. Importantly, the amplitude of the N450 conflict-related enhancement was inversely correlated with the number of prior episodes. This pattern was not found for the N2 component. These findings suggest that deficits in cognitive control increase with each depressive episode and persist after symptom remission, suggesting that depressive episodes leave a "scar" on cognitive control processes.