Working memory in social anxiety disorder: better manipulation of emotional versus neutral material in working memory

ABSTRACT

Individuals with social anxiety disorder (SAD) engage in post-event processing, a form of perseverative thinking. Given that deficits in working memory might underlie perseverative thinking, we examined working memory in SAD with a particular focus on the effects of stimulus valence. SAD (n = 31) and healthy control (n = 20) participants either maintained (forward trials) or reversed (backward trials) in working memory the order of four emotional or four neutral pictures, and we examined sorting costs, which reflect the extent to which performance deteriorated on the backward trials compared to the forward trials. Emotionality of stimuli affected performance of the two groups differently. Whereas control participants exhibited higher sorting costs for emotional stimuli compared to neutral stimuli, SAD participants exhibited the opposite pattern. Greater attention to emotional stimuli in SAD might facilitate the processing of emotional (vs. neutral) stimuli in working memory.

KEYWORDS:

• Working memory
• executive control
• social anxiety disorder
• attention

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. We also conducted analyses collapsing positive and negative stimuli together and compared sorting costs for neutral vs. emotional stimuli. Sorting costs for neutral stimuli was higher than for emotional stimuli in the SAD group, t(30) = 2.43, p = .02. In contrast, the CTL group exhibited the opposite pattern of higher sorting costs for emotional vs. neutral stimuli, t(19) = −2.83, p = .01.

2. We also conducted alternative analyses to examine whether our findings are due to high levels of depressive symptoms in the SAD group. To this effect, we conducted three separate hierarchical linear regression analyses, in which we predicted sorting costs from the LSAS scores, controlling for group, BDI scores, and the two remaining sorting costs. That is, for example, when we predicted sorting costs for negative stimuli, sorting costs for neutral and positive were included. For positive sorting costs, the LSAS scores was not a significant predictor, β = −0.15, p = .53. The LSAS scores were significant and unique predictors for negative, β = −0.44, p = .04, and neutral, β = 0.40, p = .05, sorting costs, suggesting that the current findings cannot be fully accounted for by high levels of depressive symptoms in the SAD group. The BDI scores did not emerge as significant predictors in the final models for positive, β = −0.30, p = .81, negative, β = 0.08, p = .67, and neutral, β = −0.08, p = .63.