Anxiety dissociates the adaptive functions of sensory and motor response enhancements to social threats

Abstract

Efficient detection and reaction to negative signals in the environment is essential for survival. In social situations, these signals are often ambiguous and can imply different levels of threat for the observer, thereby making their recognition susceptible to contextual cues – such as gaze direction when judging facial displays of emotion. However, the mechanisms underlying such contextual effects remain poorly understood. By computational modeling of human behavior and electrical brain activity, we demonstrate that gaze direction enhances the perceptual sensitivity to threat-signaling emotions – anger paired with direct gaze, and fear paired with averted gaze. This effect arises simultaneously in ventral face-selective and dorsal motor cortices at 200 ms following face presentation, dissociates across individuals as a function of anxiety, and does not reflect increased attention to threat-signaling emotions. These findings reveal that threat tunes neural processing in fast, selective, yet attention-independent fashion in sensory and motor systems, for different adaptive purposes.DOI: http://dx.doi.org/10.7554/eLife.10274.001

Prioritized neural processing of social threats during perceptual decision-making

“…In the present detection tasks, evidence strength corresponds to the intensity of the displayed emotion or color mask. On the basis of recent studies (El Zein et al, 2015b; Patron, Mennella, Messerotti Benvenuti, & Thayer, 2019; Wyart, Myers, & Summerfield, 2015; Wyart, Nobre, & Summerfield, 2012a), instead of computing event-related averages, we performed single-trial regressions of EEG signals against these variables. As the detection criterion (bias parameter) was different between the emotion and color tasks (see Results), we wanted to investigate differences in the neural encoding of perceptual evidence for emotion and color independently of participants’ subsequent detection of anger or violet.…”

“…(Hanks & Summerfield, 2017). Perceptual decisions based on highly relevant social signals (threat stimuli) are associated with enhanced neural encoding in sensory, associative and motor regions, from 200ms following stimulus presentation (El Zein, Wyart, & Grèzes, 2015b). Such prioritization of highly relevant social signals in sensory and motor systems was suggested to: 1. increase perceptual sensitivity to relevant features of the sensory environment, and 2. motivate rapid and adaptive behavioural responses.…”

“…Indeed, past studies used either only social stimuli – i.e. emotional displays (El Zein et al, 2015b; Pessoa & Padmala, 2005, 2007; Thielscher & Pessoa, 2007), or asked participants to categorize ambiguous stimuli as either social or non-social (Delis, Onken, Schyns, Panzeri, & Philiastides, 2016; Heekeren, Marrett, Bandettini, & Ungerleider, 2004; Philiastides, Ratcliff, & Sajda, 2006; Philiastides & Sajda, 2006, 2007; Tu, Schneck, Muraskin, & Sajda, 2017). To our knowledge, only one experiment has directly assessed the neural overlap between social and non-social decisions (Sun, Yu, & Wang, 2017).…”

“…Importantly, to tightly match social (emotion) and non-social (color) perceptual decisions, we not only equalized the amount of sensory evidence in the stimuli across tasks using morph matching, but also equalized detection sensitivity (decision difficulty) across the two stimulus dimensions (emotion and color) using an adaptive Bayesian titration procedure. We expected the neural processing of angry displays, that are behaviorally relevant for an individual’s survival, to be prioritized concomitantly in sensory and motor EEG correlates of decision-making (Conty, Dezecache, Hugueville, & Grèzes, 2012; El Zein et al, 2015b). To further isolate the specificity of social decisions, we also assessed how social anxiety influences the neural mechanisms underlying the selective processing of socially-relevant information.…”

The Decision Decoding ToolBOX (DDTBOX) – A multivariate pattern analysis toolbox for event-related potentials

“…Multivariate analysis techniques were already applied to EEG data to investigate cognition several decades ago (Gevins et al, 1979), but are currently experiencing a strong revival (for reviews and technical comments see: Bai et al, 2007;Blankertz et al, 2011;Contini et al, in press;King and Dehaene, 2014;Parra et al, 2005;Sajda et al, 2009;Stokes et al, 2015). By taking advantage of the multivariate nature of EEG signals, multivariate analysis techniques can, for example, predict the outcomes of decisions, parameters of decision models, and decision errors directly from activity patterns (e.g., Blank et al, 2013;Bode et al, 2012; 2014; Boldt and Yeung, 2015;Charles et al, 2014;Chung et al, 2015;Das et al, 2010;Parra et al, 2002;Philiastides and Sadja, 2006;Ratcliff et al, 2009;Tzovara et al, 2015;van Vugt et al, 2012; for related approaches see: El Zein et al, 2015;Wyart et al, 2012;. Others have used similar techniques to investigate visual awareness Hogendoorn and Verstraten, 2013;Fahrenfort et al, 2017), multisensory integration (Chan et al, 2017), and automatic processing of semantic features of task-irrelevant stimuli .…”