The Psychophysiological Correlates of Emotion Processing in Dysphoria

The most recent extension of the cognitive vulnerability model of depression suggests that people with mild symptoms of depression (i.e. dysphoria) will show cognitive biases primarily at early information-processing stages, while people with severe symptoms of depression (i.e. clinical depression) will show cognitive biases at late information-processing stages. To date, however, few studies have empirically explored early cognitive biases in dysphoric samples. Here, I manipulated task-relevance to functionally dissociate implicit and explicit emotional processing and used scalp electroencephalograms (EEG) to look at information-processing stages in dysphoric participants. High-density EEG was recorded during the traditional task used to study cognitive biases, the emotional Stroop task (experiment 1), and an emotional word categorization task (experiment 2). Then, in my analyses, unlike previous studies, I focused particularly on early (< 300 ms) frontal ERP effects that differentiated a group with dysphoria from a comparison group with few depression symptoms. I found that early ERP components over frontal scalp were significantly amplified in the dysphoric group, while common measures of late stage processing, such as the emotion-related late posterior positivity (LPP) and reaction time, did not differentiate groups, regardless of task. Next, to show that these effects could be replicated with non-word stimuli, I used emotional faces. Emotional faces are commonly used in ERP studies of attention and emotion, and are the most common stimuli used in neuroimaging studies of depression. As such, by using LORETA source analyses, I was able to tie my ERP findings into a wider literature. This work therefore lends support to the recent extension of the cognitive vulnerability model of depression, and contextualizes the previous cognitive bias results in the wider attention, emotion and depression literatures. This dissertation concludes with a suggestion that future studies carefully differentiate between-group and within-group effects, use different paradigms to dissociate “fast” vs. “slow” effects, and address the usefulness of early biases to predict the onset of depression through longitudinal studies.