Decades of research have investigated how our emotion and arousal interact with cognition and behavior. Nevertheless, we still have many unanswered questions that span from basic underlying mechanisms to the effects of emotion in real-life social interactions. Recent technological and methodological advances have enabled us to tackle some of these questions with novel and multi-method approaches. This symposium will highlight such methodological advances in research on emotion, arousal and cognition. In five different talks, we will bring together different methods, covering the transcutaneous vagus nerve stimulation, pupillometry, pharmacological manipulation, Virtual Reality, and machine learning in neuroimaging. Mathias Weymar will present research on the vagus nerve stimulation and discuss its effects on emotional memory. Jan Willem de Gee will present cross-species data with mice and humans, and discuss how similarly pupil-linked arousal affects decision making across different species. Ulrike Rimmele will present data on the effects of pharmacological cortisol manipulations on emotional episodic memory. Leon Kroczek will discuss how facial emotions of others alter our social behavior based on his data from Virtual Reality and a computer experiment. Finally, Michiko Sakaki will present machine-learning applications to resting-state fMRI data in understanding individual differences in emotional memory. By bringing those diverse perspectives together, this symposium will aim to provide an opportunity for lively discussions concerning their contributions to our understanding of how emotion and arousal affect human cognitive processing and behavior.
Recollection of emotional events is mediated by vagus nerve activation: Evidence from non-invasive taVNS
1University of Potsdam, Germany; 2University of Greifswald, Germany
Emotional events are better remembered than neutral ones, but the mechanisms leading to this memory advantage are not well understood in humans yet. Based on animal research, it is suggested that the memory enhancing effect of emotion is related to noradrenergic activation in the brain (via LC-arousal system), which is triggered by afferent fibers of the vagus nerve. To test the causal link between vagus nerve activation and emotional memory in humans, we applied transcutaneous vagus nerve stimulation (taVNS), a novel non-invasive stimulation technique, which targets the auricular branch of the vagus nerve. In a within-subject, 3-session design, healthy participants viewed a series of unpleasant and neutral pictures while receiving continuous taVNS or sham stimulation, and one week later performed a recognition memory task. We found that taVNS, compared to sham, improved recollection memory for emotional, but not neutral, material as reflected by behavioral and electrophysiological indices (parietal ERP Old/New effect). Furthermore, brain potentials recorded during encoding also revealed facilitated attentional discrimination between emotional and neutral scenes. Our results suggest an important modulatory influence of the vagus nerve in emotional memory formation in humans. Through its putative activation of afferent vagal projections to distinct brain systems (e.g. via the LC-NA system) taVNS may be a useful method to modulate various affective and cognitive processes, which could also be of special relevance for clinical application (e.g. as adjunct treatment option for disorders associated with altered memory functions).
Pupil dilation tracks neuromodulatory responses and predicts a reduction of choice bias across species and decision domains
Baylor College of Medicine, United States of America
Decisions are often made by accumulating ambiguous evidence over time. The brain’s arousal systems are activated during such decisions. With simultaneous pupillometry and brainstem fMRI we show that rapid dilations of the pupil track responses of a number of neuromodulatory brainstem centers involved in the control of cortical arousal state, including the noradrenergic locus coeruleus, the (partly) cholinergic basal forebrain, as well as the dopaminergic substantia nigra and ventral tegmental area. The functional impact of these pupil-linked arousal responses was a reduction of decision bias. We show that the arousal-related suppression in decision bias acts on both conservative and liberal biases, and generalizes from humans to mice, and from perceptual to memory-based decisions. Our results point to a general principle of interplay between phasic arousal and decision-making, and reveal that phasic arousal accounts for a significant component of the variability of choice behavior.
Effects of pharmacological cortisol manipulations on emotional episodic memory
University of Geneva, Schweiz
In pharmacological studies using administration of hydrocortisone to increase cortisol levels or metyrapone, to lower cortisol levels, we examined the role of the stress hormone cortisol on emotional episodic memories in humans.
Administration of the pharmacological agents varied across memory phases (encoding/consolidation vs. retrieval/reconsolidation).
Encoding under increased cortisol levels altered later memory. In contrast, encoding under suppressed cortisol levels abolished the emotional memory enhancement.
Acute administration of metyrapone decreased emotional memory recall. This decrease in recall was still evident days later when metyrapone had been washed out. At the brain level, in the placebo condition, amygdala, hippocampus and ventromedial prefrontal cortex were associated with emotional memory recall. In contrast recall of emotional memories under acute cortisol suppression did not engage these regions.
In addition, we show that metyrapone affects memory reconsolidation when it is used to suppress cortisol levels after the reactivation of a memory.
How facial emotions affect action understanding during social interactions: Insights from action recognition and experience in Virtual Reality
1Department of Psychology, Clinical Psychology and Psychotherapy, University of Regensburg; 2Department of Psychology, Cognitive Neuroscience, University of Regensburg; 3Human Computer Interaction, University of Applied Sciences in Frankfurt a. M.; 4Department of Media Informatics, University of Regensburg
Face-to-face social interaction requires fast and adaptive responses to other persons’ actions. Previous studies found that observers exploit cues such as gaze or hand preshaping to infer action intentions. Importantly, facial emotions are fundamental social cues, which might allow generating adaptive responses to actions. We implemented an interactive Virtual Reality paradigm in order to investigate the influence of facial emotions on the processing of interactive actions. Virtual agents displayed either happy or angry facial emotions while they performed appetitive and aversive actions towards the participant. Participants directly interacted with the virtual agents by reacting to these actions via hand movements. We found that angry compared to happy facial emotions increased arousal ratings, heart rate, and perceived unpleasantness. Interestingly, the latter effect was greater for appetitive compared to aversive actions, highlighting the influence of facial emotions on action processing. These results were complemented by another study using an action recognition task. Using video clips of the same agents, emotions and actions while varying action completeness, we observed an emotional bias effect. Angry compared to happy facial emotions increased participants’ likelihood to judge an action as aversive. This effect was most pronounced when video clips were short and therefore hard to recognize, but was retained for video clips of intermediate length which were easy to recognize. These findings demonstrate an influence of facial emotions on action processing during social interaction. More specifically, our data suggest that facial emotions are used to infer action intentions in order to generate fast and adaptive responses.
How effective is predicting emotional memory by resting-state functional connectivity in the brain?
1University of Reading, UK; 2University of Tübingen, Germany
There have been increased interests in using BOLD signals during rest to understand brain mechanisms underlying individual differences. Research has shown that individual differences in functional connectivity during rest are correlated with individual differences in psychopathological conditions (Veer et al., 2010) and cognitive performance in old age (Nashiro et al., 2018). Recent research has extended this approach and attempted to use individual’s functional connectivity during rest in predicting individual differences in behavioral phenotypes, such as age (Liem et al., 2017), intelligence (Dubois et al., 2018a), and personality (Dubois et al., 2018b). However, it remains unclear whether it is useful to predict individual differences in emotional memory. To address this issue, we analyzed data from 303 adults (aged between 18 and 87) from the Cambridge Centre for Ageing and Neuroscience and tested whether resting-state functional connectivity can predict emotional memory enhancement effects (i.e., enhanced memory for emotional vs. non-emotional information). We also explored predictions for episodic memory for neutral information, intelligence and age. Using data from all participants, age emerged as the only reliably predicted phenotype. Follow-up analyses done for each age group (younger adults, middle aged and older adults) revealed that resting-state connectivity significantly predicted individual differences in intelligence and episodic memory for neutral information but only for younger adults. These results suggest that resting-state connectivity would be less useful in predicting individual differences in emotional memory than other characteristics, such as age or intelligence.