Electroencephalographic theta oscillatory dynamics reveal attentional bias to angry faces

Highlights

• Angry faces elicited a shorter RT and enhanced N2pc compared to happy faces.

• Angry faces induced heightened theta oscillation as opposed to happy faces.

• Theta oscillation may represent the preferential processing of threatening faces.

Abstract

Although previous studies have reported the temporal neural dynamics (e.g., N2pc component) of threat-detection orientation for face-processing (e.g., the “anger superiority effect”), little is known about the neural oscillatory features of this phenomenon. Here, we extracted the event-related potentials and event-related oscillatory power dynamics from electroencephalographic data when participants performed a visual search task consisting of realistic angry, happy, and neutral facial expressions. We observed a shorter response time and a larger N2pc for angry faces than for happy faces, suggesting the existence of preferential processing for angry faces. Similarly, elevated theta synchronization was observed for angry faces compared to happy faces, while both angry and happy faces elicited heightened theta oscillatory activity on contralateral location of target face compared to ipsilateral location. Moreover, the theta oscillation difference negatively correlated with the amplitude of the N2pc difference. Our findings suggest that the occipital theta oscillation is engaged in the search advantage of angry facial expressions. Further, our results provide evidence that the theta oscillation may reflect the processing dynamics of threat-stimuli orientation in an ever-changing environment.

Introduction

To survive in a constantly changing environment, individuals must continuously evaluate the objects around them and promptly react (e.g., escape) to potentially life-threatening stimuli (e.g., snakes, guns, and angry faces). One theory of threat detection proposes that the characteristics of facilitated detection of threatening stimuli during evolution can guide the neural system to allocate attentional resources to these threatening stimuli; thus, the responses to threatening stimuli have become faster and more efficient [5], [11], [26].

Evidence in support of this threat detection theory shows that angry faces are detected more quickly and accurately than happy faces in cognitive tasks that involve schematic faces [6,33] and realistic human faces [25], [28]. Such preferential processing of angry faces, which is referred to as the “angry superior effect”, has been demonstrated in visual search [11], [15], visual-probe [12], [14] and attentional blink paradigms [22], [23]. Furthermore, the N2pc component of an event-related potential, which affectively mirrors the shift of attention toward task-relevant stimuli, indicates attentional target selection among distracters in visual space [18], [21]. Recent electrophysiological studies report a more pronounced N2pc upon the presentation of angry faces than happy faces [14,33]. These findings suggest a preference for the detection of threat-related emotional facial expressions.

Several studies have investigated the temporal neural dynamics of the anger superiority effect (e.g., N2pc), however little is known about the neural oscillatory features of this effect. Neural oscillatory features characterized by induced cortical responses to task-relevant stimuli are thought to provide more information about the neural activity associated with attentional, emotional, and cognitive processing compared to traditional event-related potentials (ERPs) based on evoked cortical responses [13], [19]. Theta rhythm (4–8 Hz) analysis is increasingly being used in studies on attention, face perception, and emotional processing. Event-related theta oscillation is thought to have a substantial role in the processing of orientation responses and attention [10], [17]. Dowdall, et al. [10] developed a task consisting of popout and non-popout search trials to detect the neural mechanisms that underlie efficient and inefficient visual search. Their results indicated that theta oscillation, which underlies the N2pc component, may reflect the orienting of attention toward the target during the two search tasks. Moreover, synchronized theta oscillation is considered as the fingerprint of the occipital cortex in emotional paradigms involving facial expression. Indeed, when emotional faces are presented alone, the theta synchronizations induced by threat-related facial expressions (e.g., angry and fearful faces) are more intense than those induced by happy and neutral faces [4], [24]. Notably, Sun, et al. [31] assessed the neural oscillatory responses to threatening cues (e.g., snake) in a dot-probe task and observed stronger theta-band activity to threatening cues than to pleasant cues in posterior brain regions (e.g., occipito-parietal region), implying that occipital theta synchronization is effective in discriminating the valence of emotional stimuli and mirroring the processing bias for threatening stimuli. Based on these aforementioned studies, we hypothesized that theta synchronization may be closely associated with the orientation of attention to threatening stimuli in the human occipital cortex.

Here, we reveal the temporal neural dynamics and neural oscillatory features of the anger superior effect. Compared to prior work using schematic faces [6,33], we used realistic faces to maximize ecological validity. To control for potential interference factors (e.g., stimulus redundancy), we designed a visual search task in which search arrays contained a smaller number of realistic faces (i.e., four faces), and both the target face and distractor faces were posed by the same individual. Given the anger superior phenomenon observed in previous studies, we expected an apparent attentional orientation for angry faces among neutral distractor faces compared to happy faces, as expressed by shorter visual search latencies, an enhanced N2pc component, and heightened theta oscillatory activity.