Abstract
Brain-body interactions can be studied by using directed coupling measurements of fMRI oscillations in the low (0.1–0.2 Hz) and high frequency bands (HF; 0.2–0.4 Hz). Recently, a preponderance of oscillations in the information flow between the brainstem and the prefrontal cortex at around 0.15/0.16 Hz was shown. The goal of this study was to investigate the information flow between BOLD-, respiratory-, and heart beat-to-beat interval (RRI) signals in the HF band in healthy subjects with high anxiety during fMRI examinations. A multivariate autoregressive model was concurrently applied to the BOLD signals from the middle frontal gyrus (MFG), precentral gyrus and the brainstem, as well as to respiratory and RRI signals. Causal coupling between all signals was determined using the Directed Transfer Function (DTF). We found a salience of fast respiratory waves with a period of 3.1 s (corresponding to ~ 0.32 Hz) and a highly significant (p < 0.001) top-down information-flow from BOLD oscillations in the MFG to the brainstem. Additionally, there was a significant (p < 0.01) information flow from RRI to respiratory oscillations. We speculate that brain oscillations around 0.32 Hz, triggered by nasal breathing, are projected downwards to the brainstem. Particularly interesting is the driving force of cardiac to respiratory waves with a ratio of 1:1 or 1:2. These results support the binary hierarchy model with preferred respiratory frequencies at 0.32 Hz and 0.16 Hz.
Similar content being viewed by others
Introduction
The observation of the driving force of nasal respiration on neural oscillations is known since many years1. Although many studies on the rodent brain have documented oscillations in the olfactory bulb induced by nasal breathing2,3,4,5, data from humans are lacking. These respiration-entrained rhythms in animals are global, but often overlooked, although they are dominant in the delta and theta frequency range due to the high respiration rate. Studies in patients with medically intractable epilepsy and slower respiration rhythms of ~ 0.16 to 0.32 Hz revealed nasal respiration-entrained limbic oscillations in the piriform (olfactory) cortex, as well as in the amygdala and hippocampus6,7. In contrast to rodents, in humans the spatial organization of the glomerular-layer and odor-evoked potentials are different8.
Studies on nasal versus oral breathing in humans are rare, but revealed interesting functional connectivity (FC) differences. While mouth breathing is more left lateralized, nasal breathing had a more symmetrical FC pattern9. Sensorimotor areas were most important in both types of breathing, while brainstem regions appeared more often activated with nasal breathing. A cognitive study (two-back working memory task) showed stronger brain activation in fMRI recordings and connectivity during nasal breathing than during oral breathing10. Especially, the left cerebellum, left and right inferior parietal gyrus are activated in relationship to working memory. This is significant evidence that oral breathing is inappropriate, when using brain imaging techniques in cognitive studies. One problem is the exact control of both breathing types. Nasal clips can prevent respiration via nose, but oral breathing cannot be prevented completely during nasal breathing.
The findings of Zelano et al.6 obtained with intracranial EEG recordings in patients and nasal respiration-entrained oscillations of ~ 0.16 to 0.32 Hz are of special interest for the following reasons: (i) frequencies of 0.16 Hz and 0.32 Hz are preferred breathing frequencies in the binary hierarchy model of Klimesch11, (ii) the period duration (PD) of a breathing wave of ~ 3.1 s (which corresponds to oscillations at ~ 0.32 Hz) dominates during elevated anxiety in healthy fMRI participants12 and (iii) oscillations around 0.16 Hz are very likely associated with pacemaker-like activity in the brainstem prevail in states of anxiety13. Therefore, the question arises, whether there is a relationship between respiration-entrained oscillations at ~ 0.16 Hz and ~ 0.32 Hz and anxiety processes in the limbic system.
The bi-directional information flow between respiration, cortex and brainstem can be studied directly in animal experiments, in patients with intracranial electrode arrays, and via BOLD signal recordings in normal human subjects. The calculation of the Directed Transfer Function (DTF) based on the Granger causality principle offers a way to investigate directed coupling between different physiological signals14,15. The base for DTF is a multivariate autoregressive (MVAR) model, which is fitted simultaneously to the physiological signals including BOLD recording from prefrontal cortex and brainstem. This method of processing was already successfully applied to cardiac beat-to-beat interval (RRI) time courses, respiration and BOLD data in the low frequency (LF) bands of 0.05–0.15 Hz and 0.1–0.2 Hz16,17.
The goal of this paper is to study the information flow between the prefrontal cortex (PFC), brainstem, respiration, and RRI signals in the HF band (0.2–0.4 Hz) in a group of ten selected healthy fMRI participants with high anxiety (Rassler et al.12). Important questions are: (i) Do the results of DTF calculations in the HF band allow any conclusions about nasal respiration and (ii) can respiratory oscillations at ~ 0.3 Hz be induced through slow RRI fluctuation?
Methods
Study approval
All participants gave informed written consent to the protocol of the study, which was approved by the local Ethics Committee at the University of Graz (number: GZ. 39/75/63 ex 2013/14). Our research was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.
Goal of the Study
In a group of ten healthy fMRI participants with high anxiety score the directed (causal) coupling was studied between BOLD, respiration and cardiac beat-to-beat interval (RRI) oscillations in the high frequency band 0.2–0.4 Hz.
Experimental paradigm
The experimental design consisted of four resting states (R1, R2, R3, R4) each lasting for 350 s. In each resting state the state anxiety (AS) was assessed with the state-trait anxiety and depression inventory (STADI18). Respiration and the electrocardiogram (ECG) were recorded inside the scanner. Details about the exact timing see16,17.
Participants
A total of N = 23 participants (12 female, 22 right-handed) between 19 and 34 years of age (M = 24, SD = 3.2 years) took part in the fMRI study. Participants were naïve to the purpose of the study, had no former MRI experience and were without any record of neurological or psychiatric disorders (assessed by self-report). From these 23 participants with fMRI-related anxiety already two publications are ble in Scientific Reports16,17 which focused on the two low frequency bands of 0.05–0.15 Hz and 0.1–0.2 Hz. In this paper we studied the high frequency band 0.2–0.4 Hz in a group of ten fMRI participants with the highest anxiety scores (AS). AS ranged from 10 (no anxiety) to 40 (highest) with AS = 24.6 + − 2.5. These ten participants have been subjected to an extensive wave-by-wave analyses12.
Physiological signal recording and RRI time courses
ECG from thorax and respiration with a chest belt were recorded. The sampling rate was 400 Hz. QRS detection and subsequent computation of RRI time series were performed using fMRI plug-in for EEGLAB19. For further improvement of cardiac RRI detection the Kubios HRV Premium Package20 was used. For details of the wave-by-wave analyses of breathing and RRI waves see12.
Resting state fMRI and ROI selection
Functional images were acquired with a 3 T scanner (Magneton Skyra, Siemens, Erlangen, Germany) using a multiband GE-EPI sequence21 with a simultaneous six-band acquisition with TE/TR = 34/871 ms, 52° flip angle, 2 × 2 × 2 mm3 voxel size, 66 contiguous axial slices (11 × 6), acquisition matrix of 90 × 104 and a FOV of 180 × 208 mm2. Finally, the AAL atlas22 was used to extract time courses for specified regions of interest (ROIs) in the left pre-central gyrus (PCG, ROI 1), left middle frontal gyrus (MFG, ROI 7), left insula (ROI 29), left amygdala (ROI 41), left hippocampus (ROI 37), left frontal medial orbital part (ROI 25), left brainstem (ROI 93, ROI 103) and right brainstem (ROIs 96, 98, 100).
Computing of causal coupling and statistics
The interaction between the time series was studied by means of DTF14, which allows to estimate causal coupling between signals as a function of frequency. DTF is based on the Granger causality principle23, which gives a statement on predictability for two time series. If the variance of the prediction error for the second time series is reduced by including past measurements from the first time series in the linear regression model, then the first time series can be said to cause the second time series. Granger causality principle is equivalent to the 2-channel Multivariate Autoregressive Model (MVAR) but it may be extended to an arbitrary number of channels. For this study MVAR was extended to 6 channels15. Epochs of 40 s were used for the analysis, and the bootstrap approach was employed for the statistics. More details can may be found in16,17.
Results
The group of 10 fMRI participants with high anxiety scores (AS = 24.6 ± 2.5) has been subjected to a wave-by-wave analysis, where the PDs of single breathing and cardiac RRI waves were determined12. The results of this wave-by-wave study are summarized in Table 1. The healthy subjects with high anxiety displayed a large number of slow RRI waves with PD = 10.2 s ± 0.64 (corresponds to ~ 0.1 Hz oscillations) and PD = 6.7 s ± 0.21 (corresponds to ~ 0.15 Hz oscillations) and also a large number of breathing waves with PD = 3.2 s ± 0.26 (corresponds a respiration frequency of ~ 0.31 Hz). While the RRI waves with PD of 6.7 s appeared in about 24% of the resting state duration of 320 s, the fast breathing waves appeared in about 71%.
Table 1 shows results of the phase-coupling analyses between BOLD and RRI oscillations in the 0.1–0.15 Hz band. Phase-coupling, studied in the brainstem ROIs (93, 97, 103) revealed either a positive time delay (pTD) or a negative time delay (nTD) in connection with the percentage of significant bins (sigbin%) (24,25). The former (pTD) is characteristic for neural BOLD components with RRI preceding BOLD oscillations, the latter (nTD) represents vascular BOLD components with BOLD preceding RRI oscillations. Only one participant (6R1) exhibited vascular BOLD components with origin in the baroreflex loop. In the majority of participants with pTDs between 0.8 and 2.4 s it can be assumed, that a pacemaker in brainstem was activated to manage the elevated anxiety level.
Directed coupling in the HF (0.2–0.4 Hz) band
The causal coupling analysis for the 0.2–0.4 Hz band revealed a strong information flow from middle frontal gyrus (MFG; ROI 7) to all other signals (Fig. 1). Most impressive was the coupling strength between BOLD in the MFG and the respiration signal (p < 0.001). In addition, a significant information flow from RRI to respiration signal was observed (p < 0.01). The precedence of slow RRI waves (relative to breathing waves) in the majority of subjects most likely indicate that RRI waves with oscillations at 0.16 Hz or 0.1 Hz drive breathing waves with a frequency ratio of 1:2 or 1:3 (2 or 3 breaths during one cardiac cycle). The direction of the flow is from RRI to breathing waves.
In Table 2 it can be noted the highly significant (p < 0.001) directed coupling between the BOLD signal from the MFG, respiration and BOLD oscillations in the brainstem. Such a significant flow was never observed before, neither in the LF or the 0.1–0.2 Hz bands. Therewith, it underlines both, the importance of respiratory oscillations around 0.3 Hz and the MFG as location for the downwards information flow.
BOLD oscillations during fast respiration at ~ 0.32 Hz
High anxiety is accompanied by fast respiration at ~ 0.32 Hz, a strong downwards information flow from the MFG to the brainstem, and coupling between respiration and RRI signal in the 0.2–0.4 Hz band (Fig. 1). From intracranial EEG (iEEG) recordings in patients with medically intractable epilepsy it is known that nasal breathing at frequencies between ~ 0.16 and 0.32 Hz can reveal respiration-entrained limbic oscillations in the piriform (olfactory) cortex as well as in the amygdala and hippocampus6,7. One question is, whether fast respiration at ~ 0.32 Hz in healthy humans can also elicit fast oscillations in the limbic system.
Examples of two typical subjects with high anxiety are displayed in Fig. 2A and B. The fast BOLD oscillations in the amygdala are centered at ~ 0.32 Hz and the slow oscillations in the PCG (ROI 1), MFG (ROI 7), medial frontal cortex (ROI 23) and brain stem (ROIs 93, 99) centered at ~ 0.16 Hz. While the former oscillations may be related to a downward information flow, the latter may have its origin in the brainstem pacemaker exhibiting an upward flow (Pfurtscheller et al.17). Both information flows are important, because they contribute to a power increase in the LF and HF band of the HRV, an important prerequisite for successful emotional regulation26. Note, these slow BOLD oscillations are not exactly time-locked to RRI oscillation but delayed.
BOLD oscillations during respiration at ~ 0.16 Hz
In addition to the dominating breathing waves at 0.32 Hz during high anxiety, one subject exhibited 0.16 Hz and 0.1 Hz oscillations, the former in the first half of recording and the latter in the second half. RRI peak-triggered averaged waves from the first half are displayed in Fig. 3. Most impressive are the coherent breathing and RRI waves with a period of ~ 6.3 s (corresponds to 0.16 Hz) and the distinct BOLD oscillations in the cortex, amygdala and brainstem, delayed from the breathing and RRI waves by ~ 2–3 s. This subject exhibited a significant upwards information flow from the brainstem to the cortex in the 0.1–0.2 Hz band (Pfurtscheller et al. 2022). Remarkable in Fig. 3 are the smooth slow BOLD waves in the precentral cortex and brainstem and the rippled BOLD wave in the amygdala.
Discussion
The wave-by-wave analysis revealed in all subjects, except one, a duration of breathing waves with 3.2 s + − 0.26 (mean ± SD), corresponding to a breathing frequency at ~ 0.31 Hz12. Two findings are of particular interest, (i) the dominant fast respiration at ~ 0.32 Hz and the top-down flow (Fig. 1) and (ii) the bottom-up flow at ~ 0.16 Hz (Fig. 3).
These two frequencies correspond to two preferred breathing frequencies of the binary hierarchy model11,27. The principle of this model is, that the frequency of body oscillations (respiration, heart rate) can be predicted from brain oscillations (EEG alpha and beta rhythms). The model includes three preferred breathing frequencies at 0.08 Hz, 0.16 Hz, and 0.32 Hz, exhibiting a doubling-halving frequency relationship. In this hierarchy especially the breathing frequencies at 0.32 Hz and 0.16 Hz are of interest. This model shows that synchronization processes exist in various ways and varying degrees between neural activity in peripheral neural clusters, such as the heart and stomach and brain rhythms, maintaining a general steady state during waking consciousness28.
Fast BOLD oscillations of neural and non-neural origin
The recording of BOLD oscillations with frequencies in the HF (0.2–0.4 Hz) band in the scanner, especially from amygdala (ROI 41), is a great challenge. The sampling rate of 1.15 Hz (corresponds to TR = 0.871 s) allows to record correctly sinusoidal frequency components < = 0.58 Hz. The problem is the respiratory BOLD artefact due to vessel motion (vascular BOLD), whereby areas with large blood vessels and good perfusion are affected. Such regions are localized e.g., close to the cingulum supplied by the middle cerebral artery29 and the pons with the basilar artery. Interestingly, not only slow respiratory BOLD artefacts in pons (ROI 96, 98,100) are dominant about 0.3 s prior the respiratory signal measured with chest belt (examples see Fig. 3 in30) but also fast respiratory BOLD artefacts as documented in Fig. 2A and B). This short delay between respiratory signal and vascular BOLD signal is indicated by the connection of equivalent peaks by a stippled line. The vascular BOLD signal is considered to result from a respiratory modulation of sympathoexcitatory neurons in the rostroventrolateral reticular nucleus31 causing maximal vasodilatation (maximum of the BOLD signal) immediately preceding the start of expiration and vasoconstriction (minimum of the BOLD signal) just before the start of inspiration30. The observation indicates that the BOLD artefact in the HF band can be reconstructed correctly. This, provides evidence that neural BOLD oscillations in the HF band can be studied. We speculate therefore, that fast BOLD oscillations are of neural origin and dominant in the amygdala. They are delayed relatively to the respiratory BOLD artefact through the neuro-BOLD coupling time of about 2–3 s13,32,33. The problem is that in the case of wave periods shorter than 5 s, no clear separation of vascular and neural BOLD signals is possible. Therefore, the superposition of both BOLD signals can result in a truncation of the neural BOLD response.
Breathing oscillations at ~ 0.32 Hz and top-down flow
A novel finding is the saliency of fast respiration at ~ 0.32 Hz during high anxiety. Fast respiration centered around the preferred breathing frequency at 0.32 Hz covers a broad range from 0.25 to 0.39 Hz11. Interestingly, already normal subjects while sitting in a quiet room and breathing quietly can be separated into two groups. From these, the group with higher anxiety revealed a breathing frequency of 0.28 Hz34. Both, Zelano et al.6 and Herrero et al.7 studied intracranial EEG (iEEG) in patients with medically intractable epilepsy. Zelano et al.6 controlled nasal breathing by a pneumotachometer positioned in front of the nostrils and measured mouth breathing with a chest belt. In the patients studied, a breathing rate between 0.24 and 0.36 Hz was dominant. This main finding provides strong evidence for nasal respiratory-entrained local field potential activity in human piriform cortex, amygdala, and hippocampus, whereby these fast oscillations in PFC are diminished by mouth breathing. It is assumed that such patients have a burden of anxiety, whereby the level of anxiety is unknown. A similar breathing rate around 0.31 Hz during high fMRI-related anxiety was reported in healthy subjects12.
An interesting point is the relationship between fast nasal breathing in patients and fast breathing of healthy people in the scanner during high anxiety. In the former case it is very likely that patients with medically intractable epilepsy are not free from anxiety, in the latter case the fast breathing might be one strategy or mechanism to manage high anxiety processing. From directed coupling measures there is strong evidence, that a downwards flow from the cortex to the brainstem in the 0.2–0.4 Hz band can be a marker for nasal breathing.
Breathing oscillations at ~ 0.16 Hz with bottom-up and top-down information flow
In contrast to BOLD oscillations of neural origin at frequencies around 0.32 Hz and their possible overlap with a respiratory BOLD artefact, slow neural BOLD signals at 0.16 Hz are more easily to interpret, especially if they appear coherent with respiration (Fig. 3). For the interpretation of the directed coupling data from 10 subjects shown in Fig, 1, it may be helpful to present one example with low anxiety and top-down flow from MFG in the 0.1–0.2 Hz band (Fig. 4A) and averaged BOLD waves (Fig. 4B). Remarkably are in this case the coherent BOLD oscillations at 0.16 Hz, which were found in the PFC, MFG, brainstem and limbic system (amygdala and hippocampus).
Noteworthy is the great similarity of BOLD waves in the amygdala (ROI 41) in different subjects (Figs. 3 and 4). Both figures display very similar patterns with coherent beathing and RRI waves, coherent BOLD waves in the left amygdala and left caudal pons structure (ROI 103), but opposed information flows. In case of Fig. 3, the information flow is bottom-up from the brainstem to the PFC and in case of Fig. 4, it is top-down from the PFC to the brainstem. The latter case of downwards directed coupling is a good example of respiration-entrained brain oscillations, and the former case, upwards directed coupling, can be seen as example of pacemaker-entrained brain oscillations. This information flow arrives from a pacemaker-like rhythmic source localized in reticular structures of the brainstem13,35,36. It generates rhythmic activity basically around 0.16 Hz and respiratory oscillations at the same frequency (ratio 1:1) or the double frequency (ratio 1:2). Note, these upwards coupling may suggest another possible pathway from the preBötzinger complex in the brainstem to the locus coeruleus and, moreover, via the medial thalamus and the olfactory bulb to the PFC and the limbic system37,38.
Table 1 exhibits slow RRI waves (PD = 6.7 s + − 0.21) of various appearance between 2% and 54% in all subjects with high anxiety. It is not clear, whether they are pacemaker- entrained or respiration- entrained. Motor functions such as eye or finger tracking movements can be entrained by respiration(39,40 indicating that respiration can modulate and even entrain central neural rhythms. Respiration-entrained brain rhythms are global but often overlooked and have been preferably studied in the rodent brain2,4,41,42. There is a great scientific interest in such an entrainment in humans, because the respiratory rhythm can have a direct influence on cortical oscillations such as EEG theta and gamma oscillation43. Further research on this issue is needed.
Recently, data from 28 healthy volunteers during a steady contraction task of the forearm have been recorded in a 275 channel whole head magnetoencephalography (MEG) study with visually monitored deep nasal breathing44. In this study, the corticomuscular coherence was synchronized with respiration showing a maximum over the sensorimotor cortex. Noteworthy, the dominant breathing rate was 0.14 Hz, which is close to the preferred breathing frequency of 0.16 Hz of the binary hierarchy model11.
Conclusions
During anxiety, three body oscillations with frequencies at around 0.32 Hz, 0.16 Hz and 0.1 Hz are of interest. The former two exhibit a doubling- halving frequency relationship and are predicted by the binary hierarchy model11. The latter frequency does not belong to the binary hierarchy, but may play an important role in combination with 0.16 Hz. Rassler et al.12,45 reported two center frequencies in coherent breathing and cardiac RRI waves, one at ~ 0.16 Hz and another at ~ 0.1 Hz. Preferred breathing oscillations at 0.32 Hz with a period of 3.13 s are very frequent during high anxiety processing and exhibit often doubling during one RRI wave at 0.16 Hz and triplication during one RRI waves at 0.1 Hz.
Here we suggest the hypothesis that nasal breathing is responsible for relatively slow breathing at 0.16 Hz during low anxiety. Slow breathing appears to represent a default mode in a state of alert wakefulness when nasal breathing dominates. Experience from everyday observations suggests that awake healthy persons typically prefer nasal breathing at rest. Studies on airway flow through the oral and nasal breathing routes confirmed that more than 80% of healthy subjects breathe predominantly or even exclusively through the nose at rest46,47.
In contrast, relatively fast breathing at 0.32 Hz can be observed during high anxiety. Whether breathing at 0.32 Hz is also associated with nasal breathing is not completely clear, although causal coupling revealed a significant information flow from prefrontal cortex to centers in the brainstem in the band 0.2–0.4 Hz. In contrast to this highly significant downwards information flow, an ascending flow in the band 0.1–0.2 Hz was reported by Pfurtscheller et al.17. In this latter case the source of slow rhythmic activity is very likely a pacemaker in the brainstem. Note, that pacemaker induced activity from the brainstem operates either at a center frequency of 0.16 Hz or 0.1 Hz and modulates breathing oscillations either in a 1:1 (coherent breathing and RRI waves), 1:2 or 1:3 (two or three breaths at one cardiac wave) manner.
One core function of the brain is to efficiently coordinate and regulate energetic requirements of the body. Examples are the pacemaker-entrained ~ 0.16-Hz oscillations and the preferred breathing oscillations at ~ 0.16 Hz of the binary hierarchy model of Klimesch11. Of interest are also adaptively altering physiological activities of the body (e.g. respiration and heart rate) that respond to environmental changes (e.g. changes due to claustrophobia during scanning) known as allostasis48,49,50. Possible examples of allostasis are the respiration-entrained 0.16-Hz oscillations during low anxiety, the pacemaker-entrained 0.16 Hz oscillations, and the 0.32 Hz respiration oscillations during high anxiety16,17. The importance of respiration as integral rhythm of the brain’s neural activity is highlighted in two recently published papers from Northoff’s group51,52. Respiration is less a confounder variable in fMRI data, but more an intrinsic part of the brain's entrainment and synchronization processes.
Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
References
Adrian, E. G. Olfactory reaction in the brain in the hedgehog. J. Physiol. 100, 459–473 (1942).
Biskamp, J., Bartos, M. & Sauer, J. F. Organization of prefrontal network activity by respiration-related oscillstions. Sci. Rep. 7(1), 1–11 (2016).
Heck, D. H., Kozma, R. & Kay, L. M. The rhythm of memory: how breathing shapes memory function. J Neurol. Physiol. 122, 563–571. https://doi.org/10.1152/jn.00200.2019PMID:3121534451 (2019).
Tort, A. B. L., Brankack, J. & Draguhn, A. Respiration-entrained brain rhythms are global but often overlooked. Trends Neurosci. 41(4), 186–197 (2018).
Girin, B. et al. The deep and slow breathing characterizing rest favors brain respiratory drive. Sci. Rep. 11(1), 7044 (2021).
Zelano, C. et al. Nasal respiration entrains human limbic oscillations and modulates cognitive function. J. Neurosci. 36(49), 12448–12467 (2016).
Herrero, J. L., Khuvis, S., Yeagle, E., Cerf, M. & Mehta, D. Breathing above the brain stem: Volitional control and attentional modulation in humans. J. Neurophsiol. 119, 145–159 (2018).
Lane, G., Zhou, G., Noto, T. & Zelano, C. Assessment of direct knowledge of the human olfactory system. Exp. Neurol. 329, 113304 (2020).
Jung, J. Y., Park, C. A., Lee, Y. B. & Kang, C. K. Investigation of functional connectivity differences between voluntary respirations via mouth and nose using resting state FMRI. Brain Sci. 10, 704 (2020).
Jung, J. Y. & Kang, C. K. Invstigation on the effect of oral breathing on cognitive activity using functional brain imaging. Health Care 9, 645 (2021).
Klimesch, W. The frequency architecture of brain and body oscillations: an analyses. Europ. J. Neurosci. 48, 2431–2453 (2018).
Rassler, B., Schwerdtfeger, A., Schwarz, G. & Pfurtscheller, G. Negative respiratory sinus arrhythmia (nRSA) in the MRI-scanner - a physiologic phenomenon observed during elevated anxiety in healthy persons. Physiol. Behav. 245, 113676 (2022).
Pfurtscheller, G. et al. Verification of a central pacemaker in brain stem by phase-coupling analysis between HR interval- and BOLD-oscillations in the 0.10–0.15 Hz frequency band. Front. Neurosci. 14, 922. https://doi.org/10.3389/fnins.2020.00922 (2020).
Kaminski, M. & Blinowska, K. J. A new method of the description of the information flow in the brain structures. Biol. Cybern. 65(3), 203–210 (1991).
Blinowska, K. J., Kus, R. & Kaminski, K. Granger causality and information flow in multivariate processes. Phys. Rev. E 70, 050902 (2004).
Pfurtscheller, G. et al. Processing of fMRI-related anxiety and bi-directional information flow between prefrontal cortex and brainstem. Sci. Rep. 11, 22348. https://doi.org/10.1038/s41598-021-01710-8 (2021).
Pfurtscheller, G., Blinowska, K. J., Kaminski, M., Rassler, B. & Klimesch, W. Processing of fMRI-related anxiety and information flow between brain and body revealed a preponderance of oscillations at 0.15/0.16 Hz. Sci. Rep. 12, 9117 (2022).
Spielberger, C. D., Gorssuch, R. L., Lushene, P. R., Vagg, P. R. & Jacobs, G. Manual for the State-Trait Anxiety Inventory (Consulting Psychologists Press Inc., 2009).
Niazy, R. K., Beckmann, C. F., Iannetti, G. D., Brady, J. M. & Smith, S. M. Removal of fMRI environment artifacts from EEG data using optimal basis sets. Neuroimage 28, 720–737 (2005).
Tarvainen, M. P., Niskanen, J.-P., Lipponen, J. A., Rantaho, P. O. & Karjalainen, P. A. Kubios HRV - Heart rate variability analysis software. Comput. Methods Prog. Biomed. 113(1), 210–220 (2014).
Moeller, S. et al. Multiband multislice GE-EPI at 7 Tesla, with 16-fold acceleration using partial parallel imaging with application to high spatial and temporal whole-brain fMRI. Mag. Reson. Med. 63, 1144–1153 (2010).
Tzourio-Mazoyer, N. et al. Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage 15, 273–289 (2002).
Granger, C. W. J. Investigating causal relations by econometric models and cross-spectral methods. Econometrica 37(3), 424 (1969).
Lachaux, J., Rodriguez, E., Martinerie, J. & Varela, F. J. Measuring phase synchrony in brain signals. Hum. Brain Mapp. 208, 194–208 (1999).
Pfurtscheller, G. et al. Brain-heart communication: Evidence for “central pacemaker” oscillations with a dominant frequency at 0.1 Hz in the cingulum. Clin. Neurophysiol. 128, 183–193. https://doi.org/10.1016/j.clinph.2016.10.097 (2017).
Thayer, J. F. & Lane, R. D. Claude Bernard and the heart-brain connection: further elaboration of a model of neurovisceral integration. Neurosci. Biobehav. Rev. 33, 81–88. https://doi.org/10.1016/j.neubiorev.2008.08.004 (2009).
Klimesch, W. An algorithm for the EEG frequency architecture of consciousness and brain body coupling. Front. Hum. Neurosci. 7, 766 (2013).
Young, A., Hunt, T. & Ericson, M. The slowest shared resonance: A review of electromagnetic field oscillations between central and peripheral nervous system. Front. Hum. Neurosci. 15, 833 (2022).
Birn, R. M., Diamond, J. B., Smith, M. A. & Bandettini, P. A. Separating respiratory-variation-related fluctuations from neuronal-activity-related fluctuations in fMRI. Neuroimage 31, 1536–1548 (2006).
Pfurtscheller, G. et al. “Switch-off” of respiratory sinus arrhythmia nay be associated with the activation of an oscillatory source (pcemaker) in the brainstem. Front. Physiol. 10, 939 (2019).
Haselton, J. R. & Guyenet, P. G. Central respiratory modulation of medullary sympathoexcitatory neurons in rat. Am. J. Phys. 256, R739–R750. https://doi.org/10.1152/ajpregu.1989.256.3.R739 (1969).
Mateo, C., Knutsen, P. M., Tsai, P. S., Shih, A. Y. & Kleinfeld, D. Entrainment of arteriole vasomotor fluctuations by neural activity is a basis of blood-oxygenation-level-dependent “resting-state” connectivity. Neuron 96(4), 936-948.e3 (2017).
Arthurs, O. J. & Boniface, S. How well do we understand the neural origin of the fMRI BOLD signal?. Trends Neurosci. 25, 27–31. https://doi.org/10.1016/S0166-2236(00)01995-0 (2002).
Kato, A., Takahashi, K. & Homma, I. Relationship between trait and respiratory parameters during quiet breathing in normal subjects. J. Physiol. Sci. 68, 369–376. https://doi.org/10.1007/s1256-017-0539-7 (2017).
Lambertz, M. & Langhorst, P. Simultaneous changes of rhythmic organization in brainstem neurons, respiration, cardiovascular system and EEG between 0.05 Hz and 0.5 Hz. J. Auton. Nerv. Syst. 68, 58–77 (1998).
Perlitz, V. et al. Cardiovascular rhythms in the 0.15-Hz band: common origin of identical phenomena in man and dog in the reticular formation of the brain stem?. Eur. J. Physiol. 448, 579–591. https://doi.org/10.1007/s00424-004-1291-4 (2004).
Del Negro, C. A., Funk, G. D. & Feldman, J. L. Breathing matters. Nat. Rev. Neurosci. 19, 351–367. https://doi.org/10.1038/s41583-018-0003-6 (2018).
Allen, M., Varga, S. & Heck, D. H. Respiratory rhythms of the predictive mind. Psychol. Rev. https://doi.org/10.1037/rev0000391 (2022).
Rassler, B. Mutual nervous influences between breathing and precision finger movements. Eur. J. Appl. Physiol. 81, 479–485. https://doi.org/10.1007/s004210050071 (2000).
Rassler, B. & Raabe, J. Co-ordination of breathing with rhythmic head and eye movements and with passive turnings of the body. Eur. J. Appl. Physiol. 90(125–130), 2003. https://doi.org/10.1007/s00421-003-0876-5 (2003).
Heck, D. H. et al. Breathing as a fundamental rhythm of brain function. Front. Neural Circ. 10, 115 (2017).
Bagur, S. et al. Breathing-driven prefrontal oscillations regulate maintenance of conditoned-fear evoked freezing independently of initiation. Nat. Commun. 12(1), 2605 (2021).
Tort, A. B. L., Hammer, M., Zhang, J., Brankack, J. & Draguhn, A. Temporal relations between cortical network oscillations and breathing frequency during REM sleep. J. Neurosci. 41(24), 5229–5242 (2021).
Kluger, D. S. & Gross, J. Depth and phase of respiration modulate cortico-muscular communication. Neuroimage 22, 117272 (2020).
Rassler, B., Schwerdtfeger, A., Aigner, C. S. & Pfurtscheller, G. “Switch-off” of respiratory sinus arrhythmia can occur in a minority of subjects during functional magnetic resonance imaging (fMRI). Front Physiol. https://doi.org/10.3389/fphys.2018.01688 (2018).
Niinimaa, V., Cole, P., Minitz, S. & Shephard, R. J. Oronasal distribution of respiratory airflow. Respir. Physiol. 43(1), 69–75. https://doi.org/10.1016/0034-5687(81)90089-x (1981).
Amis, T. C., O’Neill, N. & Wheatley, J. R. Oral airway flow dynamics in healthy humans. J. Physiol. 515, 293–298. https://doi.org/10.1111/j.1469-7793.1999.293ad.x (1999).
Sterling, P. Allostasis: A model of predictive regulation. Physiol. Behav. 106(1), 5–15. https://doi.org/10.1016/j.physbeh.2011.06.004 (2012).
Boyadzhieva, A., & Kayhan, E. Keeping the breath in mind: Respiration, neural oscillations, and the free energy principle. Front. Neurosci. 742 (2021).
Katsumi, Y., Theriault, J. E., Quigley, K. S. & Barrett, L. F. Allostasis as a core feature of hierarchical gradients in the human brain. Netw. Neurosci. 6(4), 1010–1031 (2022).
Goheen, J., Anderson, J. A. E., Zhang, J. & Norhoff, G. From lung to brain – respiration modulates neural and mental activity. Avaiable under PsyArXiv (2022).
Zhang, J. & Northoff, G. Beyond noise to function: Reframing the global brain activity and its dynamic topography. Commun. Biol. 5(1), 1350 (2022).
Acknowledgements
We like to thank David Fink for support in data acquisition, Clemens Brunner for support in data processing and Andreas Schwerdtfeger for evaluations of state anxiety (all of them from University of Graz).
Author information
Authors and Affiliations
Contributions
G.P. and K.B. conceived and designed study. M.K. performed the processing of the data and statistics. G.P. wrote the manuscript. B.R. G.S. and W.K. reviewed the final version.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Pfurtscheller, G., Kaminski, M., J.Blinowska, K. et al. Respiration-entrained brain oscillations in healthy fMRI participants with high anxiety. Sci Rep 13, 2380 (2023). https://doi.org/10.1038/s41598-023-29482-3
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-29482-3
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.