Effects of task load, spatial attention, and trait anxiety on neuronal responses to fearful and neutral faces

Abstract

Facial emotional expressions constitute a significant part of communication as they transfer non-verbal signals to others. Especially for expressions signaling threat or danger, distinct early and mid-latency event-related potentials (ERPs) modulations are reported (e.g., see Mühlberger et al., 2009;Schupp et al., 2004;Yoon et al., 2016), even if facial emotion remains task-irrelevant, suggesting high automaticity of emotion discrimination by specific ERP components Smith, 2012).

Potentiated early neural responses to fearful faces are not driven by specific face parts

“…Besides the N170, also the EPN was larger for fearful compared to neutral faces in accordance with various previous studies 27 , 44 , 59 , 87 , 88 , confirming that the EPN differentiates between fearful and neutral faces in a paradigm that does neither direct attention away from the stimuli 27 , nor requires attentional focus on the emotional expression 70 . Classification maps reveal that especially the left-hemispheric EPN is sensitive to facial features and that this sensitivity is restricted to low spatial frequencies.…”

“…One important trait related to altered threat detection and perception is trait anxiety, which has been proposed to lead to hypervigilant processing of threat-related stimuli [19][20][21][22] . However, while early studies reported increased early ERPs for fearful faces in high trait-anxious participants [23][24][25] , other studies found no effect 26,27 or even attenuated differential ERPs 28,29 . Thus, although theoretical arguments (e.g., 21,23 ) strongly suggest a relation between trait anxiety and early neural responses to threat, findings are inconclusive.…”

“…Fearful expressions have been shown to increase further this ERP component 6 . Notably, increased N170 amplitudes by emotional expressions are not explained by low-level information 8 , 40 , 43 and are highly resistant to various attentional manipulations 27 , 37 , 40 , 44 – 47 . Both holistic accounts 47 – 50 and accounts assuming a role of specific features, such as mouth 51 – 54 or eyes 53 – 56 have been proposed to explain N170 modulations by faces in general and fearful facial expressions in particular.…”

“…Furthermore, these studies required participants to identify the emotional expression on each trial, creating specific attentional demands for participants. Task relevance has been shown to affect ERPs, especially the EPN and later components, to facial expressions and the differences between them 15 , 27 , 37 , 70 . Although task relevance is a valid setting in which to study effects of facial expressions on ERPs, we suggest that task irrelevant faces yield a more default-like and thus ecologically more valid insight into emotional face processing.…”

Fearful faces straight ahead or in the periphery: Early neuronal responses independently of trait anxiety.

“…P1 amplitude modulations might reflect recurrent effects of the amygdala on activations in visual areas, which have been associated with the P1 across different methodological approaches (intracranial recordings, see Méndez-Bértolo et al, 2016; lesion studies, see Mielke et al, 2022; Rotshtein et al, 2010; combined EEG-fMRI measures, see Müller-Bardorff et al, 2018). Of note, given the notoriously lower reliability of P1 emotional modulations (Schindler & Bublatzky, 2020), several previous studies showed no significant P1 emotion effects for central (e.g., see Itier & Neath-Tavares, 2017; Neath-Tavares & Itier, 2016; Schindler, Richter, et al, 2022) and peripheral locations for fearful or angry faces (Rigoulot et al, 2012; Schindler, Busch, et al, 2022) or peripheral simple negative pictures (Carretié et al, 2017, 2020). This discrepancy might be explained by specific features of our study, including the relatively large EEG sample, the sustained attention to peripheral locations, and the use of monetary reward.…”