From the heart to the brain: Central nervous effects of cardioafferent signals
|Zusammenfassung der Sitzung|
Peripheral bodily processes are conveyed to the brain via several mechanisms promoting psychological and behavioral adaptations. Besides humoral pathways, the impact of neural projections originating from receptors located within the cardiovascular system is currently debated. Accumulating evidence indicates that cardiac activity influences central-nervous processes mediating perception, cognition and emotion. This symposium will cover recent neuroscientific research into such heart-brain interactions ranging from basic perception to psychopathology. Michael Gaebler (Leipzig) will present evidence for a modality-specific cardiac phase bias indicating that increased cardioafferent traffic during systole inhibits somatosensory perception while promoting active visual sampling. Mauro Larra (Dortmund) will present behavioral and EEG studies showing that cardiac activity modulates sensorimotor and cognitive processes underlying stimulus-response-compatibility. Although central-nervous effects of phasic variations in cardioafferent traffic can be revealed by analyzing EEG data time-locked to heartbeats, electrical fields propagated from the heart pose a challenge and Stefan Arnau (Dortmund) will discuss methodological approaches to deal with the cardiac field artifact. Conscious and unconscious perception of heartbeats may also contribute to self-related cognitions and emotions. Aleksandra Herman (Warsaw) will identify neural correlates of heart-focused and tactile-focused attention in a study applying fMRI and a novel heartbeat-detection task. André Schulz (Luxemburg) will focus on psychopathological consequences related to the perception of cardioafferent signals and present new results on heart-beat-evoked potentials and interoception in somatic symptoms. Together, these contributions elucidate how cardioafferent signals influence brain activity as well as the way we feel, perceive and interact with our environment.
Somatosensory perception and active visual sampling vary across the cardiac cycle
1Max-Planck-Institut für Kognitions- und Neurowissenschaften, Deutschland; 2Charité – Universitätsmedizin Berlin; 3University of Warsaw; 4Humboldt-Universität zu Berlin
The processing of signals in the environment is influenced by spontaneous activity not only in the brain but also in the heart. During each cardiac cycle, when the ventricles contract and eject blood into the arteries (i.e., systole), stretch-sensitive baroreceptors signal blood pressure changes to the brainstem and (sub)cortical brain regions. During diastole, the ventricles relax and baroreceptors remain inactive.
In two independent studies with healthy adults (total n=70, 26±4 years), passively presented near-threshold somatosensory stimuli were more likely to be perceived when they were presented during later (i.e., diastole) compared to earlier phases (i.e., systole) of the cardiac cycle. Using simultaneous EEG, we found higher heartbeat-evoked potential amplitudes over somatosensory electrodes to be associated with lower detection rates and lower somatosensory-evoked potential amplitudes.
Instead of passively receiving signals that coincide with certain neural or cardiac states, animals actively explore their environment and interact with it. Movement-based sensory sampling (i.e., active perception) can align sensory inflow to time points that are optimal for information processing. This was shown for eye movements, which are preferentially generated at specific phases of oscillatory activity in the brain and the heart (i.e., systole). In study 3, healthy adults (n=47, 26±4 years) actively prompted the onset of briefly presented pictures, which they had to memorise, by self-paced button press. Pictures were preferentially displayed during systole, while recognition memory was not affected by the heartbeat.
In conclusion, how we perceive the world and engage with it also depends on the state of our heart.
How cardioafferent traffic modulates the processing of response conflicts
Neural projections originating from peripheral receptors within the cardiovascular system have been shown to impact on central nervous processing of exteroceptive stimuli. Cardioafferent signals are processed mainly in the insula and anterior cingulate cortices, areas that are also critically involved in conflict and error monitoring. Indeed, recent evidence suggests that these processes may be influenced by cardioafferent signaling. Here, we will present electrophysiological and behavioral data on how cardioafferent traffic modulates the processing of response conflicts. In a series of experiments, lateralized tactile, auditory and visual stimuli were presented synchronized to the ECG during either high (systole) or low (diastole) cardioafferent traffic. Participants were asked to respond to the stimuli with either spatially compatible or incompatible manual button presses. Along with behavioral parameters, event-related lateralizations (ipsi-contra differences, ERLs) were analyzed time-locked to stimulus and response onset in the EEG. We found that spatially compatible responses to auditory but not tactile stimuli were speeded during systole, whereas incompatible reactions were slowed. Moreover, this was also reflected in the EEG by a distinct modulation of compatibility effects in ERLs at frontocentral recording sites during systole but not diastole. Finally, in a visual Simon task we found speeded reactions and increased ERLs during systole, however, this effect was not moderated by spatial stimulus-response compatibility. In sum, these results suggest a role of cardioafferent traffic in the processing of spatial stimulus-response conflicts but the effects also seem to depend on stimulus modality and task parameters.
Removing the cardiac field artefact from the EEG using neural network non-linear regression
Leibniz Research Centre for Working Environment and Human Factors (IfADo), Deutschland
When measuring EEG, the electric field generated by cardiac activity is captured by the scalp electrodes as the cardiac field artifact (CFA). In most experimental setups CFA related variance is unsystematic and therefore diminished by averaging procedures when parameterizing the EEG signal. This is not the case, however, when the variance of interest happens to be time-locked to cardiac activity. In studies investigating heartbeat evoked potentials or in studies deploying stimulation at certain phases of the cardiac cycle, the CFA might become a dominant feature and a confound in averaged measures like the ERP or time frequency decompositions. We present a non-linear regression approach deploying neural networks to remove the CFA from the EEG signal. The core concept of the approach is to use R-peak centered episodes that are free of systematic variance elicited by experimental stimulation to train neural network models to predict the EEG signal from the ECG signal. In a second step, these trained models can then be used to predict the EEG from ECG during episodes with experimental stimulation (i.e. during trials). We show for lateralized visual stimuli presented time-locked to the ECG, that removing these predictions from the signal effectively removes the CFA from the signal without affecting stimulus-related variance.
Neural divergence and convergence for attention to and detection of interoceptive and somatosensory stimuli
Nencki Institute of Experimental Biology, Poland
Body awareness is constructed by signals originating from within and outside the body. How do these apparently divergent signals converge? We developed a signal detection task to study the neural convergence and divergence of interoceptive and somatosensory signals. Participants focused on either cardiac or tactile events and reported their pres- ence or absence. Beyond some evidence of divergence, we observed a robust overlap in the pattern of activation evoked across both conditions in frontal areas including the insular cortex, as well as parietal and occipital areas, and for both attention and detection of these signals. Psycho-physiological interaction analysis revealed that right insular cortex connectivity was modulated by the conscious detection of cardiac compared to somatosensory sensations, with greater connectivity to occipito-parietal regions when attending to cardiac signals. Our findings speak in favour of the inherent convergence of bodily-related signals and move beyond the apparent antagonism between exteroception and interoception.
Enhanced responsiveness of heartbeat-evoked potentials to attention focused on heartbeats in high somatic symptom distress: a neurophysiological explanation for medically-unexplained symptoms
1Clinical Psychophysiology Laboratory, Department of Behavioural and Cognitive Sciences, Universität Luxemburg, Esch-sur-Alzette, Luxemburg; 2Arbeitseinheit Klinische Kinder- und Jugendpsychologie, Fakultät für Psychologie, Ruhr-Universität Bochum, Bochum, Deutschland; 3Abteilung für Klinische Psychologie, Psychotherapie und Experimentelle Psychopathologie, Johannes Gutenberg-Universität Mainz, Mainz, Deutschland
The psychobiological mechanisms underlying somatic symptom generation remain unclear. In the perception filter model, it is posited that (1.) enhanced bodily signal transmission and (2.) decreased CNS filter function promote (3.) increased perception physical symptoms. As there is a scarcity of empirical evidence to support this model, we addressed this issue in the current project. In Study 1, we identified extreme groups of high (HSR; n=29) and low symptom reporters (LSR; n=29). In Study 2, we recruited 23 patients with somatic symptom disorder (SSD), 24 clinical control patients with major depression (MDD) and 25 healthy control individuals. Heart rate (HR), heart rate variability (HRV) and diurnal cortisol release were assessed as indicators of (1.) signal transmission. Heartbeat-evoked potentials (HEPs) served as indicators of (2.) CNS filter functions. Interoceptive accuracy (IAc) in heartbeat perception tasks may reflect (3.) perception of physical signals. Individuals with high symptom distress showed (1.) higher resting HR and lower RMSSD HRV, but no differences in cortisol release. In contrast to our expectations, (2.) higher HEP reactivity to attention focused on heartbeats reflect higher filter function in individuals with higher symptom distress. Finally, we did not observe differences in (3.) IAc between groups. The model assumptions could only be partially confirmed: SSD may be associated with (1.) stronger bodily signals, but (2.) higher CNS filter function. As more attentional resources are mobilized to process heartbeats, but perception accuracy remains unchanged (3.), this overspill could be responsible for detecting minor bodily changes contributing to somatic symptom generation.