Abstract
Negative emotions increase self-focused attention, a core feature of depression and anxiety-related disorders. However, the cognitive mechanisms associated with the tendency to self-focus—and, conversely, with the ability to voluntarily disengage attentional resources from the self and direct them towards the external environment—remain poorly understood. Thus, this study aimed to examine whether a specific cognitive control mechanism that directs attention between self-generated thoughts and external information might moderate the relationship between dysphoria and maladaptive self-referential thinking. Results showed that dysphoria increases the frequency of rumination, self-blame, and catastrophising, especially for participants who have more difficulty in switching from self-generated thoughts to information provided by the environment. These results shed new light on the cognitive mechanisms underlying maladaptive self-referential thinking associated with dysphoria. More specifically, this specific cognitive mechanism might play a key role in the maintenance or amplification of a depressed mood.
ACKNOWLEDGEMENTS
Many thanks to J. Hislair and P. Mayer for their help in the data collection. We also thank the anonymous reviewers for their relevant comments on an earlier version of the manuscript.
Notes
1Note that difficulties in shifting from internal to external information moderate the relationships between dysphoria and maladaptive self-referential thinking, as reflected by a factor that collapsed these dimensions, as well as by each score taken separately.
2To respond to a referee's comment, we computed an additional regression analysis, including interaction effects between working memory and stimulus-independent to stimulus-oriented switch cost, gender and stimulus-independent to stimulus-oriented switch cost, and the three-way interaction working memory × stimulus-independent to stimulus-oriented switch cost × gender. The results of this analysis showed that only dysphoria is a significant predictor of maladaptive self-referential thinking, β = 0.49, t(64) = 4.74, p < .00001, but that neither working memory nor gender interacts with shifting abilities to predict maladaptive self-referential thinking (all ps > .10). However, the sample size might not be large enough to highlight three-way interactions.