Electrical neuroimaging reveals content-specific effects of threat in primary visual cortex and fronto-parietal attentional networks
Introduction
Attention can bias sensory processing early on following stimulus onset. In monkeys, as well as in humans, top-down attentional gating effects have been shown as early as in the thalamic nuclei (Fischer and Whitney, 2012, McAlonan et al., 2008, O'Connor et al., 2002) and V1 (Schwartz et al., 2005). Yet, affective states provide powerful motivational drives that can influence attentional deployment, sometimes conflicting with top-down goal setting. Sustained anxious anticipation (Davis et al., 2010) of unpredictable and uncontrollable bodily harm (uncued delivery of electric shocks, for example) has consistently been implicated in humans and animals in hypervigilant threat monitoring (Alvarez et al., 2011, Davis and Whalen, 2001, Somerville et al., 2010), which is related to augmented sensory vigilance in order to facilitate threat detection. Critically, although anxious hypervigilance can foster an effective monitoring of the environment, it comes at a price. Stress and anxiety have been shown to induce neural plasticity in key regions such as the hippocampus, the amygdala and the prefrontal cortex (Leuner and Shors, 2013, McEwen et al., 2012), altering cognitive functions, such as emotion regulation, attentional control (Arnsten, 2009, Bishop, 2007, Eysenck et al., 2007, Plessow et al., 2011) and goal-directed stimulus processing (Moser et al., 2005, Shackman et al., 2011).
However, outside the laboratory, threatening events are not restricted to the imminence of potential bodily harm: uncertainty, social stressors, or upsetting visual scenes are also able to trigger anxious responses, implicating activations in the extended amygdala, similarly to physical threats (Grupe et al., 2013, Yassa et al., 2012). Unlike bodily harm, these psychosocial strains have been suggested to narrow the attentional scope, resulting in decreased early visual responses to irrelevant sensory information (Easterbrook, 1959, Rossi and Pourtois, 2012a, Schmitz et al., 2009), without systematically affecting goal-directed behavior. By comparison, sustained anxiety related to the anticipation of uncontrollable physical harm seems mainly to impair attentional control functions, in favor of a bottom-up (ventral) attentional system, which might mediate hypervigilance (thus, enhanced responses to task-irrelevant, but potentially threatening information, see also Bishop et al., 2004, Choi et al., 2012, Cornwell et al., 2011, Pourtois et al., 2013).
Furthermore, negative affect can be elicited by increasing task difficulty alone, in the absence of a direct mood induction (Nummenmaa and Niemi, 2004). Hence, the typical narrowing of the attentional focus associated with high load tasks (Lavie, 2005, Rauss et al., 2009, Schwartz et al., 2005) might actually be conflated by an uncontrolled increase of negative affect, given that this state has also been related to narrowed attention (Easterbrook, 1959).
Thus, although different forms of anticipatory anxiety and distress seem to impinge on stimulus processing in dissociable ways, no study to date has directly compared their differential effects on specific attention control processes. This may be explained by the challenges posed by bringing up these different affective states in the laboratory and comparing them in a systematic way. For example, matching negative affect intensity between different types of strain (physical vs. psychosocial) appears especially challenging in between-subjects experimental designs. On the other hand, having the same participants experience these different strains in a within-subjects design may lead to uncontrolled carryover, with the residual effects of one specific state possibly contaminating the subsequent one.
To overcome these problems, we used in this study a novel experimental design enabling to compare in the same participants the effects of two different types of anticipatory anxiety (physical vs. psychosocial) on the electrophysiological markers of attention control, while minimizing systematic carryover effects. In this paper, we operationalized anticipatory anxiety, or state anxiety, as a state of sustained tension in the anticipation of the possible encounter with a negative event which is not imminent, but looming (Davis et al., 2010). We continuously measured the skin conductance levels (SCLs) to monitor peripheral arousal during the experience of state anxiety, in order to be able to model the physiological response corresponding to the stressor anticipation against the normal habituation curves (measured in a condition in which no stressors were expected). Two types of anticipatory anxiety inductions (physical vs. psychosocial) were chosen given their similarity to published procedures in the literature. For each of them, we could then formulate a clear prediction regarding the specificity of its effects on attentional processes: while the threat of bodily harm would primarily impair goal-directed processing (see for example Shackman et al., 2011), by contrast, an “internal” psychological stressor should be accompanied by a narrowing of the attention focus (e.g., Rossi and Pourtois, 2012a, Schmitz et al., 2009). In agreement with previous reports, we formally operationalized attention control as the ability to maintain efficient goal-directed processing (indexed by the P300 ERP component, see Shackman et al., 2011), as well as the ability to filter out irrelevant stimuli in early sensory cortex (C1 component, indexing attentional gating in V1, see Rauss et al., 2011). We compared effects of anticipatory anxiety driven by physical vs. psychosocial threat on attention control brain processes to a third condition, consisting of enhanced perceptual load, given its well-known effects on both goal-directed processing within the fronto-parietal network (e.g., Lavie, 2005, Schwartz et al., 2005) and early attentional filtering in the primary visual cortex (see Rauss et al., 2009, Rauss et al., 2011). Importantly, by measuring self-report distress and autonomic arousal responses, we could also assess whether a possible increase in negative affect would arise in this condition (albeit of lower magnitude compared to the two state anxiety conditions), an element which has typically been overlooked in earlier studies investigating effects of load on attention selection.
At the behavioral level, we expected high load to slow down reaction times and decrease accuracy for target detection, compared to low load. Since psychosocial threat does not seem to systematically affect overt behavior (no effects on target detection or discrimination were reported in studies investigating the narrowing of attention during negative affect induced by feedbacks or upsetting picture presentation; see Moriya and Nittono, 2011, Rossi and Pourtois, 2012a, Rossi and Pourtois, 2013, Schmitz et al., 2009), we therefore surmised that task performance would not be influenced by this form of anticipatory anxiety. Concerning the effects of bodily threat, a recent review by Robinson et al. (2013) highlighted the lack of consistency in behavioral costs during the anticipation of unpredictable noxious stimuli (threat of mild electric shocks). Nevertheless, two recent electrophysiological studies using the threat of bodily harm (Moser et al., 2005, Shackman et al., 2011) did not report differential effects at the behavioral level between the safe vs. threat condition. Accordingly, we reckoned that bodily threat would not lead to an impaired behavioral performance in this experiment.
However, based on existing dissociations in the literature (e.g., peripheral distractor processing seems to be either decreased or increased under stress depending on situational factors, see Choi et al., 2012 and Schmitz et al., 2009), we predicted that the two types of strains (physical vs. psychosocial) would have dissociable effects on attention control at the electrophysiological level. Hypervigilance and reduced goal-directed processing were hypothesized to be related to the threat of bodily harm, selectively (Choi et al., 2012, Moser et al., 2005, Shackman et al., 2011). Sensory hypervigilance would primarily be translated in maintained or increased early responses to irrelevant information in V1 (C1 component, Weymar et al., 2013) accompanied by an attenuation of goal-directed processing as measured in amplitude of the target-locked P300 (Moser et al., 2005, Shackman et al., 2011). By comparison, we surmised psychosocial threat to narrow the attentional focus around fixation, and therefore selectively increase filtering of peripheral (irrelevant) information in V1 (C1 component). At the same time, goal-directed processing should remain relatively unaffected, as previously reported (Moriya and Nittono, 2011, Rossi and Pourtois, 2012a, Schmitz et al., 2009).
To corroborate the assumption of systematic changes in the fronto-parietal network during goal directed processing (P300 effect) as well as in V1 during the early filtering of irrelevant information (C1 effect) as a function of these two strains, we estimated the intra-cerebral sources of these two ERP components using a distributed inverse solution (standardized Low Resolution Brain Electromagnetic Tomography, sLORETA).
Section snippets
Participants
Twenty-six right handed undergraduates participated in the study (mean age = 20.4 years, SD = 2.2 years, 7 males). Participants had normal or corrected-to-normal vision, were unaware of the purpose of the study and declared no history of psychiatric or neurological disorders, nor the use of psychoactive medication. The study protocol was conducted in accordance with the Declaration of Helsinki and approved by the local ethics committee.
Stimuli and task
The paradigm was adapted from Rossi and Pourtois (2012a).
Self report and SCL/SCR results
Prior to the start of the experimental session, participants' reported state anxiety levels were comparable to published norms (STAI-Spre M = 35, SD = 7.4, range 20–52). At the end of the experimental session, state anxiety was significantly increased (STAI-Spost M = 40, SD = 8.5, range 23–54; T25 = 4.51, P = .0001).
To gain a better understanding of the dynamical changes in the participants' state anxiety levels, we tested whether our induction procedures were both successful in enhancing distress.
Discussion
A wide range of contexts perceived as challenging, uncertain or dangerous have the potential to elicit anticipatory anxiety responses, entailing long-lasting activations in the extended amygdala (Davis and Whalen, 2001, Somerville et al., 2010), and increased autonomic arousal. However, the question arises as to whether different threat-inducing situations, although sharing some common neurobiological mechanisms, actually map onto the same realm or, instead, correspond to different situations,
Acknowledgments
This work was supported by a grant from the European Research Council (Starting Grant #200758 to GP), by the Special Research Funds at Ghent University (BOF Grant #05Z01708 to GP and #BOF13/PDO/095 to VR) and by the Belgian Science Policy, Interuniversity Attraction Poles program (P7/11).
References (71)
- et al.
Phasic and sustained fear in humans elicits distinct patterns of brain activity
NeuroImage
(2011) Neurocognitive mechanisms of anxiety: an integrative account
Trends Cogn. Sci.
(2007)- et al.
Medial prefrontal cortex damage affects physiological and psychological stress responses differently in men and women
Psychoneuroendocrinology
(2010) - et al.
Impact of state anxiety on the interaction between threat monitoring and cognition
NeuroImage
(2012) - et al.
Anxiety overrides the blocking effects of high perceptual load on amygdala reactivity to threat-related distractors
Neuropsychologia
(2011) - et al.
Perceptual load interacts with involuntary attention at early processing stages: Event-related potential studies
NeuroImage
(2009) - et al.
A standardized boundary element method volume conductor model
Clin. Neurophysiol.
(2002) - et al.
D1 receptors in prefrontal cells and circuits
Brain Res. Rev.
(2000) - et al.
A new method for off-line removal of ocular artifact
Electroencephalogr. Clin. Neurophysiol.
(1983) - et al.
Repetition and the brain: neural models of stimulus-specific effects
Trends Cogn. Sci.
(2006)
Age effects on the P300 potential and the corresponding fMRI BOLD-signal
NeuroImage
10/20, 10/10, and 10/5 systems revisited: their validity as relative head-surface-based positioning systems
NeuroImage
Levels of appraisal: a medial prefrontal role in high-level appraisal of emotional material
NeuroImage
What does polarity inversion of extrastriate activity tell us about striate contributions to the early VEP? A comment on Ales et al. (2010)
NeuroImage
Influence of task difficulty on the features of event-related potential during visual oddball task
Neurosci. Lett.
Distracted and confused?: selective attention under load
Trends Cogn. Sci.
Stress, anxiety, and dendritic spines: what are the connections?
Neuroscience
Stress and anxiety: structural plasticity and epigenetic regulation as a consequence of stress
Neuropharmacology
Effect of mood states on the breadth of spatial attentional focus: an event-related potential study
Neuropsychologia
Effects of perceptual learning on primary visual cortex activity in humans
Vis. Res.
Brain mechanisms for emotional influences on perception and attention: what is magic and what is not
Biol. Psychol.
State-dependent attention modulation of human primary visual cortex: a high density ERP study
NeuroImage
Orienting and reorienting: the locus coeruleus mediates cognition through arousal
Neuron
Multiple synergistic effects of emotion and memory on proactive processes leading to scene recognition
NeuroImage
Localization bias and spatial resolution of adaptive and non-adaptive spatial filters for MEG source reconstruction
NeuroImage
Human bed nucleus of the stria terminalis indexes hypervigilant threat monitoring
Biol. Psychiatry
Functional MRI of the amygdala and bed nucleus of the stria terminalis during conditions of uncertainty in generalized anxiety disorder
J. Psychiatr. Res.
Stress signalling pathways that impair prefrontal cortex structure and function
Nat. Rev. Neurosci.
Trait anxiety and impoverished prefrontal control of attention
Nat. Neurosci.
State anxiety modulation of the amygdala response to unattended threat-related stimuli
J. Neurosci.
Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: the BIS/BAS scales
J. Pers. Soc. Psychol.
Control of goal-directed and stimulus-driven attention in the brain
Nat. Rev. Neurosci.
Where and when the anterior cingulate cortex modulates attentional response: combined fMRI and ERP evidence
J. Cogn. Neurosci.
The amygdala: vigilance and emotion
Mol. Psychiatry
Phasic vs sustained fear in rats and humans: role of the extended amygdala in fear vs anxiety
Neuropsychopharmacology
Cited by (21)
Effects of perceptual and working memory load on brain responses to task-irrelevant stimuli: Review and implications for future research
2022, Neuroscience and Biobehavioral ReviewsCitation Excerpt :ERP studies that exhibit a high temporal resolution are designated to fill this gap, as early parts of the C1 component, the earliest cortical ERP response to visual stimuli, are based on V1 activity. However, findings of C1 effects are inconsistent, where some studies support the view that high load might reduce C1 amplitudes (Rauss et al., 2009; Rossi and Pourtois, 2012, 2014), partially in a complex interaction with spatial attention effects (e.g., see Acunzo et al., 2019; Fu et al., 2008, 2009). Nevertheless, other studies instead support the notion of an absence of C1 modulations by load (Ding et al., 2014; Wolf et al., 2021) or report even increased C1 amplitudes during high load (Rauss et al., 2012).
Exploration of Exogenous Attention to Disgust and Fear Pictures with Different Spatial Frequencies Through Event-related Potentials
2022, NeuroscienceCitation Excerpt :To our knowledge, the present study is the first to examine C1 amplitude responses to LSF and intact emotional visual stimuli as distractors in an exogenous attention task. It had already been reported that attention can modulate the amplitude of C1 according to parameters such as attentional load, spatial location or sensory modality (Rauss et al., 2009; 2011; 2012; Rossi and Pourtois, 2014). Our findings suggest that the spatial frequency of the distractor may also modulate the amplitude of C1.
Affect-biased attention and predictive processing
2020, CognitionExamining the role of feedback in TMS-induced visual suppression: A cautionary tale
2019, Consciousness and CognitionImplicit guidance of attention: The priority state space framework
2018, CortexCitation Excerpt :Although they cannot speak to the influence of the amygdala, studies using ERPs have provided further evidence that affective/motivational modulation of visual cortex can occur rapidly and early in the visual stream. The C1 component, which is observed between 50 and 90 msec after stimulus onset and is thought to reflect activity of striate cortex, has been found to reflect prioritization of negatively conditioned stimuli (Stolarova, Keil, & Moratti, 2006), threatening stimuli (Rossi & Pourtois, 2014), and stimuli associated with the loss of reward (Rossi et al., 2017). Just as the visual cortex is modulated by fronto-parietal networks, V1 activation has been found to reflect patterns associated with biased competition for fearful relative to neutral facial expressions (West, Anderson, Ferber, & Pratt, 2011).