Clinically significant states of depression and anxiety may be associated with negative biases in cognition and judgment. Previous studies have suggested that normal mood states may also produce mood-congruent cognitive biases. We theorize that the mood states of depression and anxiety are mechanisms through which the brain regulates its arousal, hedonic expectancy, and attention. By examining dense array event-related brain potential (ERP) measures as subjects evaluate daily life events, we observed evidence that mood-congruent expectancies can be observed as early as 300 ms into the evaluation process. Additional research has shown that, in addition to the widespread ERP effects of language manipulations, there are focal electrical changes over medial frontal cortex reflecting general attentional mechanisms, such as when subjects orient to a novel event or detect they have made an error. Because these medial frontal effects appear to reflect activity in the limbic cortex at the base of frontal motor pathways, a reasonable hypothesis is that emotionally significant evaluative decisions will also engage medial frontal activity. Recordings as subjects rated themselves with trait adjectives confirmed that the initial discrimination between good and bad traits was observed in medial frontal cortex, substantially before more widespread (P300/N400) effects were seen. These findings suggest that systematic experiments may clarify the self-regulatory functions of the limbic regions of the frontal lobe. These frontolimbic networks are implicated in depression and anxiety by both lesion and neuroimaging evidence. For the next project period, we propose a research program that will integrate the findings on mood-congruent cognitive biases with the findings on frontal lobe mechanisms of self-monitoring. By coupling advanced measures of brain electrical responses with systematic psychological experiments, we may be able to clarify the normal mechanisms of affective self-regulation that may become exaggerated in mood disorders.
Showing the most recent 10 out of 18 publications
