Edited by: Maurizio Codispoti, University of Bologna, Italy
Reviewed by: Cali Bartholomeusz, University of Melbourne, Australia; Shi-Min Chen, China University of Mining and Technology, China
This article was submitted to Emotion Science, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
To examine the interaction of working memory (WM) type with emotional interference in trait anxiety, event-related potentials were measured in a combined WM and emotional task. Participants completed a delayed matching-to-sample task of WM, and emotional pictures were presented during the maintenance interval. The results indicated that negative affect interfered with spatial WM; task-related changes in amplitude were observed in the late positive potential (LPP) and slow waves in both the high and low anxiety groups. We also found an interaction among WM type, emotion, and trait anxiety such that participants with high levels of trait anxiety showed an opposite neural response to verbal and spatial WM tasks compared with individuals with low trait anxiety during the sustained brain activity involved in processing negative or neutral pictures in the delay phase. Our results increase our understanding of the influence of emotions on recognition and the vulnerability of those with trait anxiety to emotional stimuli.
► This study examined the effects of emotional distraction on working memory in individuals with trait anxiety.
► Participants with high trait anxiety showed an opposite neural response to verbal and spatial working memory tasks compared with participants with low trait anxiety.
► Negative distraction interfered with spatial working memory.
Cognitive models of anxiety suggest that impaired cognitive control plays a critical role in the development and maintenance of anxiety (Song et al.,
Within the context of cognitive control, emotional interference is the emotionally salient stimuli that may potentially impair cognitive task (e.g., perception and WM) and can compromise the ability to complete tasks requiring cognitive control (Song et al.,
Anxiety is always accompanied by changes in cognitive processing, and the effects of anxiety on cognitive performance may be mediated by their effects on WM (Eysenck and Calvo,
WM system is composed of three advanced cognitive operations: the phonological loop dealing with verbal information, the visuospatial sketchpad for processing non-verbal visual and spatial information, and the central executive for control. The central executive acts as more of an attention system—particularly in maintaining task goals and reducing interference from distraction (Moran,
Previous studies have examined the influence of emotional interference on WM. Imaging studies with a delayed matching-to-sample task demonstrated that negative emotional disturbance presented during the delay period hindered performance (Dolcos and McCarthy,
Prior research has demonstrated that distracting tasks may reduce emotion-related physiological responses, such as the event-related (ERP) components of late positive potential (LPP) and slow waves, which are sensitive to affective stimuli. Early components of LPP have been interpreted as responses to attentional capture, recognition, and stimulus evaluation (Donchin and Coles,
The aim of the present study was to examine the interaction of WM type with emotional interference in trait anxiety subjects with the ERP method. Based on existing research reviewed above, we hypothesized that the effects of anxiety on the neural correlates of WM are attributable to a specific component function of WM. Second, the influence of anxiety on the neural correlation of WM attributes to the different valences of the emotional distractors.
Participants gave written informed consent to participate in the study. And the written informed consent was approved by the local ethics committee.
Three hundred nineteen undergraduate students completed the Chinese version of the Trait Anxiety Inventory (TAI) (Spielberger et al.,
The emotional stimuli were 120 pictures consisting of 60 aversive pictures and 60 neutral pictures, which were selected from the International Affective Picture System (Lang et al.,
The WM materials were 24 letters from the Latin alphabet. To ensure that the physical characteristics of the two types of stimuli are completely identical, the stimulus sets for the verbal and spatial task were the same (Li et al.,
The participants were seated in an electronically isolated, sound- and light-attenuated room and viewed a computer monitor from a distance of 75 cm.
First, a white fixation point “+” appeared in the center of the black background for 400–600 ms. Three uppercase letters then appeared at anywhere around the “+” for 2,500 ms. The letter stimuli occupied 3–5° of visual angle on the visual midline. The participants were instructed to remember the three letters, followed by a white “+” for 1,000 ms. After the presentation of a picture (neutral or negative) for 750 ms, a white “+” appeared for 1,000 ms. Finally, a small letter appeared at 1 of the 12 points corresponding to the clock (
The sequence of events in a trial. The picture was neutral or negative.
Participants completed a short training session consisted of 12 stimuli, followed by two formal sessions. Only one kind of emotional picture appeared in one session. Sessions were separated by a 5-min interval. In order to avoid emotional disturbance, the neutral session appeared before the aversive session. Each session was composed of two blocks, one verbal and one spatial, appearing in random order. Each block included 60 trials, resulting in a total of 240 trials, which appeared completely randomized. There was a brief rest when participants finished 30 trials.
Electroencephalograms (EEGs) were recorded from 64 scalp electrodes located in standard 10/20 electrode positions embedded in an elastic cap recording device (NeuroScan version 4.3 system). All electrodes were referenced to the M2 (right mastoid) and then re-referenced offline to the average of M1 (left mastoid) and M2. EEGs were recorded with a 0.01–100-Hz bandpass filter and 1,000-Hz sampling rate. Electrode impedances were always kept below 5 kΩ. Vertical electro-oculogram (EOG) recording electrodes were positioned above and below the left eye, and horizontal EOG recording electrodes were positioned at the outer canthi of both eyes. Each epoch was filtered with a 24-Hz low-pass filter. Trials with various artifacts were rejected using a criterion of ±100 μV. The ERPs were averaged for trials with correct responses.
The EEG was segmented for each trial beginning 100 ms before to 1,000 ms after the picture onset. Based on previous studies (Hajcak et al.,
Mean error rates and reaction times (RTs) were entered into a 2 × 2 × 2 mixed analysis of variance (ANOVA), with task type (spatial/verbal) and valence (negative/neutral) as within-subject factors and group (LA/HA) as between-subject factor. Besides, all electrophysiological data were analyzed by repeated-measures ANOVA also including electrode (FC/CP/P/PO) and laterality (left/midline/right) as within-subject factors. Greenhouse–Geisser adjustments to the degrees of freedom were performed.
ANOVA results revealed a main effect of emotion [
Basic descriptive statistics of ACC and reaction time (RT) in both groups.
Verbal | Neutral | 86.35 ± 13.39% | 84.81 ± 15.91% | 803.98 ± 125 | 821.34 ± 114.95 |
Negative | 94.2 ± 4.4% | 93.76 ± 4.36% | 759.54 ± 98.37 | 759.67 ± 74.06 | |
Spatial | Neutral | 86.8 ± 9.62% | 86.1 ± 6.2% | 717.56 ± 101.85 | 756.33 ± 112.77 |
Negative | 89.85 ± 6.27% | 88.24 ± 5.16% | 679.01 ± 83.6 | 703.51 ± 100.03 |
ANOVA results revealed a main effect of task type [
The ANOVA conducted on early LPP (
Grand-average waveforms in the LA and HA groups under the four conditions.
The task type × emotion interaction,
The task type × emotion × group interaction was significant,
Regarding late LPP amplitude (
The task type × emotion × group interaction was significant,
The main effect of laterality was significant,
The interaction effect of task type × laterality × group was significant,
The ANOVA conducted on slow wave showed a main effect of emotional state [
We found a marginally significant task type × electrode × group interaction effect,
The task type × emotion interaction was marginally significant,
Topographies of difference waves formed by subtracting spatial trial ERPs from verbal trial ERPs during 296–356, 452–512, and 660–760 ms after picture onset.
This study examined the effect of emotional interference on visual WM. The main aim was the interaction of WM type with emotional interference in trait anxiety.
Our findings demonstrate differential modulation of emotional interference in WM in trait anxiety. We report that participants with high trait anxiety showed a different neural response to verbal and spatial WM tasks. Compared with the low anxiety group, this reaction pattern appeared only in the high anxiety group. This provides direct electrophysiological measures of the disruptive effects of negative emotions and anxiety on WM. In other words, anxiety combined with emotion has an effect on WM. It is not only emotions that affect WM. This modulation depends not only on WM type but also on the value of emotional disturbance. Interestingly, the interference effect of emotional pictures on spatial WM in high trait anxiety shows opposite pattern.
We initially tested for the interference effect and found greater susceptibility to distraction due to interference by negative stimuli in spatial than in verbal WM. All participants made more errors and had longer RTs to the probe for spatial compared with verbal material during the maintenance delay under the negative-emotion conditions. The behavioral data showed that interference by negative emotions impaired spatial WM. A number of studies have found poorer spatial WM performance during interference by negative affect (Dolcos and McCarthy,
Also, task performance was no different between the LA and HA groups. Previous similar study also observed no impaired performance in the HA group (Zhang et al.,
The findings suggest that the neural responses involved in the processes underpinning sustained brain activation during the delay phase of verbal and spatial WM tasks may be opposite in individuals with high trait anxiety, depending on whether they are responding to negative or to neutral pictures. In contrast, individuals with LA may not have opposite neural responses. Specifically, participants with high trait anxiety had higher early LPP amplitudes in response to neutral pictures during the maintenance phase of the spatial WM task relative to the verbal WM task. In contrast, participants with high trait anxiety had higher earlier LPP amplitudes in response to negative pictures during the maintenance phase of the verbal WM task relative to the spatial WM task.
Why was HA associated with opposite neural responses during sustained brain activation during the delay phase of verbal and spatial WM tasks involving negative vs. neutral affective pictures? This question can be answered in terms of LPP and the processing mechanisms underpinning spatial WM. The early components of LPPs represented attentional capture, recognition, and stimulus evaluation (Donchin and Coles,
Our ERP results indicated that the amplitudes of the early LPPs elicited by the spatial WM task were larger than those elicited by the verbal WM task in the HA group during the maintenance phase following the neutral pictures. This supports the assumption that attentional resources are required for spatial WM but not necessarily for verbal WM (Li et al.,
In contrast, after seeing negative pictures, the amplitude of the early LPP of the HA group was higher for the verbal compared with the spatial WM task. Donchin and Coles (
The results revealed such an effect in that the late LPP and slow-wave components also showed task-related amplitude changes in the HA and LA groups. The LPP extends throughout the entire duration of picture presentation, indexing increased sustained attention to emotional stimuli (Foti et al.,
Individuals with high trait anxiety demonstrate affective bias toward aversive or threat-related stimuli (Wilson and MacLeod,
Our behavioral and ERP results show HA individuals demonstrating disparate neural processing yet equivalent performance to LA individuals, indicating that impaired task performance is not a characteristic of HA individuals. Instead, HA individuals tend to be “strategic” in that they avoid mistakes in completing the task. The real-world implications of our results are that impaired task performance is not the only evidence to judge whether an individual is highly anxious, and EEG data should also be combined to judge.
Our findings may contribute to the comprehension of WM in trait anxiety. On the one hand, our findings demonstrate the generality of impaired cognitive control in HA individuals, not only in attentional control, but also in the inhibition of negative interference from WM. Our results extend prior observations of inhibition control failures at the stage of maintenance (Qi et al.,
The present study has some limitations. A limitation of the WM task was the use of letters for both the spatial and verbal WM conditions. This is because during the spatial condition it would have been more difficult to inhibit the automatic human nature of reading the visual stimuli than it would be to inhibit spatial location (and this extra cognitive load/attentional resources may partially explain the difference in RTs between the two conditions) and the higher early LPP amplitude during spatial WM task compared to the verbal WM task in the HA group.
In conclusion, our findings demonstrate that negative emotional distractors have differential effects on neural attentional processing for individuals with high vs. low trait anxiety, suggestive of possible compensatory/resilience mechanisms at work to allow comparative task scores in individuals with high trait anxiety. Our results suggest that such mechanisms have important effects on the performance of cognitive tasks and that the kind of effect depends on the type of the task (verbal or spatial), on the level of trait anxiety (higher levels of trait anxiety are associated with more interference), and on the type of emotional interference (neutral or negative). Our results contribute to our understanding of the influence of emotions on recognition performance and trait anxiety, which are associated with increased susceptibility to emotional disruptions.
The datasets presented in this article are not readily available because the data also forms part of the ongoing study. Requests to access the datasets should be directed to Huifang yang,
The studies involving human participants were reviewed and approved by the Ethics Committee of Lingnan Normal University. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from all participants for the publication of any potentially identifiable data included in this article.
HY: experiment design. JL: data collection. XZ: major revision of the manuscript. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The Supplementary Material for this article can be found online at: