Attentional bias, implicit memory, and general slowing in depression

Biases of visual attention to and implicit memory for emotional language were compared in patients with naturally occurring depression and nondepressed individuals. It was hypothesized that the self-referential quality of emotional (negative or positive) words would interfere with speed of color-naming, suggesting differential allocation of attention to emotional self-referential words. This was expected to occur differentially between the diagnostic groups, with depressed subjects demonstrating selective attention to negative words and nondepressed subjects demonstrating selective attention to positive words. It was further hypothesized that differential attention to emotional words, if discovered in the color-naming task, would result in a similar bias of verbal memory. Words to which subjects selectively attended on the basis of their emotional valence and/or self-referential qualities should be more quickly accessed via implicit processes from memory, as demonstrated in enhanced repetition priming effects in a subsequent lexical decision-making task. Subjects provided their own ratings of the degree to which each word was self-descriptive after performing the attention and memory tasks, thus allowing the influence of self-reference to be examined as a function of each subject's personal view of him or her self. The results indicated that, although depressed subjects selectively attended to negative emotional words, nondepressed subjects did not selectively attend to positive emotional words. Self-reference did not significantly influence attention in either group. Furthermore, despite the finding of intact (although slower) access to implicit memory traces of words seen previously by the depressed group, there was no enhancement of repetition priming of negative words. The finding of automatic selective attention to negative words in depressed individuals represents both a quantitative and a qualitative difference in cognitive processing compared to nondepressed persons. However, the pattern of repetition priming indicates a purely quantitative difference. The strong, reliable correlation of nondepressed with depressed subjects' mean reaction times (which were consistently proportionally slower in all lexical decision task conditions) suggests that general slowing accounts for the group differences in lexical access speed. The possible influences of current mood and arousal on attention and memory processes in depression are discussed.