Respiration aligns perception with neural excitability

Article activity feed

1. 

   Version published to 10.7554/elife.70907 on eLife

   Dec 14, 2021
2. 

   eLife

   Nov 29, 2021

   Author Response:

   Reviewer #1 (Public Review):

   The main finding - that the moment-to-moment relationship between excitability and perception is coupled to the body's slower respiratory oscillation - is novel, interesting, and important for advancing our understanding of how the brain-body system works as a whole. The experiment is simple and elegant, and the authors strike the right level of making the most of the data without doing too much and obscuring the main findings. The primary weakness, in my opinion, is the inability to distinguish between the possibility that respiration modulates excitability and the possibility that respiration modulates something boring like signal-to-noise ratio. In terms of conclusions, I thought the authors stuck pretty well to the data. The one place where the conclusions felt a little bold was in …

   Author Response:

   Reviewer #1 (Public Review):

   The main finding - that the moment-to-moment relationship between excitability and perception is coupled to the body's slower respiratory oscillation - is novel, interesting, and important for advancing our understanding of how the brain-body system works as a whole. The experiment is simple and elegant, and the authors strike the right level of making the most of the data without doing too much and obscuring the main findings. The primary weakness, in my opinion, is the inability to distinguish between the possibility that respiration modulates excitability and the possibility that respiration modulates something boring like signal-to-noise ratio. In terms of conclusions, I thought the authors stuck pretty well to the data. The one place where the conclusions felt a little bold was in terms of the respiration <> alpha <> behavior relationship, where it felt the authors had already made up their minds re: causality. I agree that it probably makes more sense for respiration to influence something about the brain than vice versa, and the background presented in the Intro/Discussion supports this. However, the analysis only tells us that the behavioral performance was modulated by both alpha and respiration (and their interaction, but this is no way causal). Overall, it will be necessary to differentiate the current interpretation from the possibility that breathing and alpha are two unrelated time courses that influence behavior at the same time (and even interact in how they influence behavior, but just not interact with each other), and I do not believe the phase-amplitude coupling analysis is sufficient for this.

   We thank the reviewer for their positive and constructive evaluation of our work.

   Reviewer #2 (Public Review):

   Kluger and colleagues investigated the influence of respiration on visual sensory perception in a near-threshold task and argue that the detected correlation between respiration phase and detection precision is liked to alpha power, which in turn is modulated by the phase of respiration. The experiments involved detecting a low-contrast visual stimulus to the left or right of a fixation point with contrast settings adjusted via an adaptive staircase approach to reach a desired 60% hit rate, resulting in an observed hit rate of 54%. The main findings are that mutual information between the discrete outcome of hit-or- miss and the continuous contrast variable is significantly increased when respiration phase is considered as well. Furthermore, results show that neuronal alpha oscillation power is modulated in phase with respiration and that perception accuracy is correlated with alpha power. Time resolved correlation analysis aligned on respiration phase shows that this correlation peaks during inspiration around the same phase where the psychometric function for the visual detection task reaches a minimum. The experimental design and data analysis seem solid but there are several concerns regarding the novelty of the findings and the interpretation of the results.

   Major concerns: The finding that visual perception is modulated by the respiration cycle is not new (see e.g. Flexman et al. 1974 or Zelano et al. 2016).

   There are multiple studies going back decades that show alpha oscillation power to be modulated by breathing (e.g. Stancák et al., 1993, Bing-Canar et al. 2016). Also, as the authors acknowledge, it is well-established that alpha power correlates with neuronal excitability and perception threshold. What seems to be new in this study is the use of a linear mixed effect model to analyze the relationship between alpha power, respiration phase and perception accuracy. However, the results mostly seem to confirm previous findings.

   Thank you for giving us the opportunity to clarify our approach and the conceptual novelty it provides. First, not at all do we claim that our study is the first to demonstrate respiration-related alpha changes. Not only do we prominently cite the work by Zelano and colleagues (JNeuro, 2016) in the Introduction and Discussion sections, we also have previous work from our own lab demonstrating these effects (see Kluger & Gross, PLoS Biol 2021). Second, the reviewer’s comment that ‘the results mostly seem to confirm previous findings’ unfortunately appears to frame a critical proof-of-concept as a lack of novelty: In order for us to claim a triadic relationship between respiration, excitability, and behaviour, it is paramount to first demonstrate that assumptions about pairwise relations (such as respiration <> alpha power and alpha power <> behaviour) are supported, which of course means replicating known results in our data. Third, in order to evaluate the novelty of our present study, it is crucial to consider its core aim, which was to characterise how automatic respiration is related to lowest-level perception by means of respiration-induced modulation of neural oscillations. At this point, we respectfully disagree with the reviewer’s assessment of our results being mostly replicative, as the references they provide differ from our approach in various key aspects: The classic study by Flexman and colleagues (1974) merely differentiates between inspiration and expiration, critically without accounting for the asymmetry between the two respiratory phases. Zelano and colleagues (2016) did not investigate visual perception at all, but instead asked participants to categorise emotional face stimuli (termed ‘emotion recognition task’). Stancák and colleagues (1993) did not investigate automatic, but paced breathing, which involves continuous, conscious top-down control of one’s breathing rhythm - a demand that is not comparable to automatic, natural breathing we investigate here. The same is true for any kind of respiratory intervention or training like the ‘mindfulness-of-breathing exercise’ employed in the study by Bing-Canar and colleagues (2016). Once again, the oscillatory changes reported by the authors are not induced by automatic breathing, but instead reflect the outcome of a conscious manipulation of the breathing rhythm. In highlighting the key differences between previous studies and our approach, we do hope to have dispelled the reviewer’s initial concern regarding the novelty of our findings.

   Magnetoencephalography captures broad band neuronal activity including gamma frequencies. As the authors show (Fig. 4) and other studies have shown, the power of neuronal oscillations across multiple frequency bands is modulated by respiration phase. Gamma and beta oscillations have been implicated in sensory processing as well. Support for the author's hypothesis that the perception threshold modulation with respiration is due to alpha power modulation would be strengthened if they could show that the power of oscillations in other frequency bands are not or only weakly linked to perception accuracy.

   We thank the reviewer for their well-justified suggestion to extend the spectral scope of our analyses to include other frequency bands. In response to their comment, we have recomputed our analysis pipeline for the frequency range between 2 - 70Hz. While the whole analysis and results are described in a new Supplementary Text and Supplementary Figures (see below), we outline key findings here.

   In keeping with the structure of our main analyses, we first computed cluster-corrected whole-scalp topographies for delta, theta, alpha, beta, and gamma bands for hits vs misses over time intervals 1s prior to stimulus presentation:

   

   Fig. S4 | Band-specific topographies over time. Whole-scalp topographic distribution of normalised pre- and peristimulus power differences between hits and misses, separately for each frequency band. Channels with significant differences in the respective band are marked (cluster-corrected within the respective time frame). Related to Fig. 3.

   Compared to the clear parieto-occipital topography of prestimulus alpha modulations, delta and theta effects were prominently shifted to anterior sensors, which renders their involvement in low-level visual processing highly unlikely. No significant effects were observed in the gamma range. In contrast, beta-band modulations were closest to the alpha effects in their topography, covering parietal as well as occipital sites. Although the size of normalised effects were markedly smaller in the beta band (compared to alpha frequencies, cf. colour scaling), the topographic distribution of prestimulus modulations as well as the spectral proximity of the two bands prompted further investigation of beta involvement. To this end, we computed the instantaneous correlation between individual beta power (over the respiration cycle) and respiratory phase, analogous to our main analysis shown in Fig. 4c. Consistent with the TFR analysis shown above, no significant correlation between oscillatory power and respiration time courses were found for delta, theta, and gamma bands. For the beta band, however, we found a significant correlation during the inspiratory phase, similar to the alpha correlation described in the main text (and shown for comparison in the new Supplementary Fig. S5):

   

   Fig. S5 | Instantaneous correlation of beta power and perceptual sensitivity. Group-level correlation between individual beta and PsychF threshold courses (averaged between 14 - 30 Hz) with significant phase vector (length of seven time points) marked by dark grey dots (cluster-corrected). Correlation time course of the alpha band (see Fig. 4c) shown for reference in light grey. Related to Fig. 4.

   While both alpha and beta power were correlated to the breathing signal during the inspiratory phase, the correlation time courses suggested that there might be differential effects in both frequency bands, as indicated by the phase shift visible in Supplementary Fig S5. Therefore, we finally recomputed the LMEM visualised in Fig. 4 with an additional factor for beta power. In this extended model, significant effects were found for both alpha (t(1790) = 3.27, p < .001) and beta power (t(1790) = 4.83, p < .001). Beta showed significant interactions with the sine of the respiratory signal (t(1790) = -3.52, p < .001) as well as with alpha power (t(1790) = -4.63, p < .001). Comparing the LMEM to the previous model which only contained alpha power (along with respiratory sine and cosine) confirmed the significant contribution of beta power in explaining PsychF threshold variation by means of a theoretical likelihood ratio test (χ²(4) = 60.43, p < .001). Overall, we thus found beta power to be i) significantly modulated by respiration (see Fig 1), ii) significantly suppressed over parieto-occipital sensors for hits vs misses (see Fig. S4), and iii) significantly contribute to variations in PsychF threshold (see Fig S5). Collectively, these findings suggest differential roles of alpha and beta power, which we discuss in the main text as well as in the Supplementary Text:

   “Whole-scalp control analyses across all frequency bands demonstrated that this topographical pattern was unique to alpha and beta prestimulus power (see Supplementary Text 1 and Fig. S4).”

   “Control analyses across all frequency bands yielded a significant instantaneous correlation between PsychF threshold and beta power as well, albeit at a slightly later phase (see Fig. S5). No significant correlations were found for the remaining frequency bands.”

   “Accordingly, one recent study proposed that the alpha rhythm shapes the strength of neural stimulus representations by modulating excitability (Iemi et al., 2021). Previous work by Michalareas and colleagues (2016) as well as our own data (see Supplementary Material) point towards an interactions between alpha and beta bands, as beta oscillations have very recently been implicated in mediating top-down signals from the frontal eye field (FEF) that modulate excitability in the visual cortex during spatial attention (Veniero et al., 2021). Our findings suggest that this top-down signalling is modulated across the respiration cycle in a way that changes behavioural performance.”

   In the discussion the authors speculate that respiration locked modulation of alpha power and associated neuronal excitability could be based on the modulation of blood CO2 levels. Most recent studies of respiratory modulation of brain activity have demonstrated significant differences between nasal and oral breathing, with nasal breathing (through activation of the olfactory bulb) typically resulting in a stronger influence of respiration on neuronal activity and behavioral performance than oral breathing. The authors only tested nasal breathing. If blood CO2 fluctuations are indeed responsible for the observed effect, there should be no difference in outcome between nasal and oral breathing. Comparing the two conditions would thus provide interesting additional information about the possible underlying mechanisms.

   We appreciate the reviewer’s well-justified remarks regarding the differential effects for nasal and oral breathing and their implications on underlying mechanisms such as CO2. In revising the present as well as other manuscripts, it has become evident that fluctuations of CO2 alone (and, as we previously discussed, related changes in pH) cannot possibly explain the effects we and others are observing. Therefore, the revised manuscript no longer discusses CO2 as a potential mechanism. We have removed the corresponding paragraph and instead refer to the distinction between nasal and oral breathing to strengthen the argument for OB-induced cross-frequency coupling:

   “As outlined in the introduction, there is broad consensus that cross-frequency coupling (Canolty and Knight, 2010; Jensen and Colgin, 2007) plays a central role in translating respiratory to neural rhythms: Respiration entrains neural activity within the olfactory tract via mechanoreceptors, after which the phase of this infraslow rhythm is coupled to the amplitude of faster oscillations (see Fontanini and Bower, 2006; Ito et al., 2014). While this mechanism is difficult to investigate directly in humans, converging evidence for the importance of bulbar rhythms comes from animal bulbectomy studies (Ito et al., 2014) and the fact that respiration-related changes in both oscillatory power and behaviour dissipate during oral breathing (Zelano et al., 2016; Perl et al., 2019). Thus, rhythmic nasal respiration conceivably aligns rhythmic brain activity across the brain, which in turn influences behaviour. In our present paradigm, transient phases of heightened excitability would then be explained by decreased inhibitory influence on neural signalling within the visual cortex, leading to increased postsynaptic gain and higher detection rates. Given that the breathing act is under voluntary control, the question then becomes to what extent respiration may be actively used to synchronise information sampling with phasic states of heightened excitability.”

   Reviewer #3 (Public Review):

   The topic is timely, the study is well-designed, and the work has been performed in a highly competent manner. The authors relate three variables: respiration, alpha power and perceptual performance, constituting a link between somatic and neuronal physiology and cognition. A particular strength is the temporal resolution of respiration effects on cognition (continuous analysis of the respiration cycle). Furthermore, results are well contextualized by very comprehensively written introduction and discussion sections (which, nevertheless, could be slightly shortened).

   We do appreciate the reviewer’s positive evaluation of our manuscript and are thankful for their constructive remarks. We respond to their comments in detail below and have shortened the Discussion section in response to one of the reviewer’s remarks (kindly see points 1.1 and 2 below).

   I have three points of criticism, all meant in a constructive way:

   1. I wonder whether the authors could have gone one step further in the analysis of causal mechanisms, rather than correlations. The analysis of timing (Fig. 4d) and the last sentence of the abstract suggest that they imagine a causal role of respiratory feedback on cognitive performance, mediated via coordination of brain activity (in the specific case, by increasing excitability in visual areas). This could be made more explicit by appropriate experiments and data analysis:

   1.1. Manipulating the input signal: former studies suggest that nasal respiration is crucial for effects on brain oscillations and/or performance (e.g. Yanovsky et al., 2014; Zelano et al., 2016). Thus, the causal inference could be easily checked by comparing nasal versus oral respiration, without changing gas- and pH-parameters of activity of brainstem centers. >Admittedly, this experiment may add significant work to the present data which, by themselves, are already very strong.

   We thank the reviewer for their insightful comment regarding the question of causality. We acknowledge that our interpretation should have been phrased a little more cautiously. Therefore, we have rephrased corresponding paragraphs at various instances throughout the manuscript (kindly see below). Particular under current circumstances, we further appreciate the reviewer’s concern regarding the acquisition of additional data for a direct comparison of nasal vs oral breathing. Their comment is of course entirely valid and we were eager to address it, especially since it relates to CO2- and/or pH-related mechanisms of RMBOs we previously discussed. In light of the reviewer’s comments (also see their related comment #2 below) and convincing evidence from both animal and human studies that already compared nasal and oral breathing, we no longer feel that changes in CO2 provide a reasonable explanation for respiration-related oscillatory and behavioural effects we observed here. Consequently, we have removed the corresponding paragraph from the Discussion section which now reads as follows:

   “As outlined in the introduction, there is broad consensus that cross-frequency coupling (Canolty and Knight, 2010; Jensen and Colgin, 2007) plays a central role in translating respiratory to neural rhythms: Respiration entrains neural activity within the olfactory tract via mechanoreceptors, after which the phase of this infraslow rhythm is coupled to the amplitude of faster oscillations (see Fontanini and Bower, 2006; Ito et al., 2014). While this mechanism is difficult to investigate directly in humans, converging evidence for the importance of bulbar rhythms comes from animal bulbectomy studies (Ito et al., 2014) and the fact that respiration-related changes in both oscillatory power and behaviour dissipate during oral breathing (Zelano et al., 2016; Perl et al., 2019). Thus, rhythmic nasal respiration conceivably aligns rhythmic brain activity across the brain, which in turn influences behaviour. In our present paradigm, transient phases of heightened excitability would then be explained by decreased inhibitory influence on neural signalling within the visual cortex, leading to increased postsynaptic gain and higher detection rates. Given that the breathing 17 act is under voluntary control, the question then becomes to what extent respiration may be actively used to synchronise information sampling with phasic states of heightened excitability.”

   1.2. Temporal relations: The authors show that respiration-induced alpha modulation precedes behavioral modulation (Fig. 4d and related results text). Again, this finding suggests a causal influence of respiration on performance, mediated by alpha suppression (see results, lines 318-320). Could the data be directly tested for causality (e.g. by applying Granger causality, dynamic causal modelling or other methods)? If this is difficult, the question of causality should at least be discussed more explicitly.

   We appreciate the reviewer’s constructive criticism and their suggestion to employ causal analyses. While we agree that the overall pattern of results strongly suggests a causal cascade of respiration -> excitability -> perception, our interpretation with regard to a dynamic mechanism was probably overly strong. Unfortunately, it is indeed difficult to use directional analyses like Granger causality or DCM on the current data, since these methods quantify the relationship between two time series. They would not allow us to investigate the triad of respiration, alpha power, and behaviour, as we have discrete responses (i.e., single events) instead of a continuous behavioural measure. In fact, we are currently preparing a directional analysis of respiration-brain coupling (in resting-state data without a behavioural component) for an upcoming manuscript. In response to the reviewer’s remarks, we have toned down our interpretation throughout the manuscript and explicitly discuss the question of causality in the Discussion section of the revised manuscript:

   “The bootstrapping procedure yielded a confidence interval of [-33.17 -29.25] degrees for the peak effect of alpha power. While these results strongly suggest that respiration-alpha coupling temporally precedes behavioural consequences, they do not provide sufficient evidence for a strict causal interpretation (see Discussion)”

   “Rigorous future work is needed to investigate potentially causal effects of respiration-brain coupling on behaviour, e.g. by means of directed connectivity within task-related networks. A second promising line of research considers top-down respiratory modulation as a function of stimulus characteristics (such as predictability). This would grant fundamental insights into whether respiration is actively adapted to optimise sensory sampling in different contexts, as suggested by the animal literature.”

   1. At various instances, the authors suggest that respiration-induced changes in pH may be responsible for the changes in cortical excitability which, in turn, affect behavioral performance. In the discussion, they quote respective literature (lines 406-418). I glanced through the quoted papers by Feldman, Chesler, Lee, Dulla and Gourine - as far as I could see none of them suggests that the cyclic process of respiration induces significant cyclic shifts of pH in the brain parenchyma (if at all, this may occur in specialized chemosensory neurons in the brainstem). Moreover, recent real-time measurements by Zhang et al. (Chem. Sci 12:7369-7376) do also not reveal such cyclic changes in the cortex. Finally, translating oscillatory extracellular pH changes (if existent) into changes in inhibitory efficacy would require some time, potentially inducing delays and variance onto the cyclic changes at the network level. I feel that the evidence for the proposed mechanism is not sufficient, notwithstanding that it is a valid hypothesis. Please check and correct the interpretation of the cited literature if necessary.

   We acknowledge the reviewer’s caution regarding our suggestion of pH involvement, which is closely related to their previous comment (kindly see 1.1 above). As the reviewer mentions themselves, there are several studies demonstrating an absence of both neural and behavioural modulations for oral (vs nasal) breathing. These reports provide direct evidence against a mechanism driven by changes in CO2 and/or pH, which would be identical for nasal and oral breathing. Moreover, a second valid criticism is the uncertain temporal delay introduced by the (hypothetical) translation of pH changes into neural signals, which would most likely be incompatible with the ‘online’ (i.e., within-cycle) effects we report here. Therefore, as outlined in our response above, we have removed the pH-related suggestions from the Discussion section.

   1. Finally, some illustrations should be presented in a clearer way for those not familiar with the specifics of MEG analysis.

   We appreciate the reviewer’s suggestions regarding the clarity of our manuscript.

   Read the original source
3. 

   eLife

   Nov 29, 2021

   Evaluation Summary:

   Kluger and colleagues investigated the influence of respiration on visual sensory perception in a near-threshold task and argue that the detected correlation between respiration phase and detection precision is liked to alpha power, which in turn is modulated by the phase of respiration. The main finding - that the moment-to-moment relationship between excitability and perception is coupled to the body's slower respiratory oscillation - poses a potentially important advance for advancing our understanding of how the brain-body system works as a whole.

   (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #3 agreed to share their name with the authors.)

   Read the original source
4. 

   eLife

   Nov 29, 2021

   Reviewer #1 (Public Review):

   The main finding - that the moment-to-moment relationship between excitability and perception is coupled to the body's slower respiratory oscillation - is novel, interesting, and important for advancing our understanding of how the brain-body system works as a whole. The experiment is simple and elegant, and the authors strike the right level of making the most of the data without doing too much and obscuring the main findings. The primary weakness, in my opinion, is the inability to distinguish between the possibility that respiration modulates excitability and the possibility that respiration modulates something boring like signal-to-noise ratio. In terms of conclusions, I thought the authors stuck pretty well to the data. The one place where the conclusions felt a little bold was in terms of the …

   Reviewer #1 (Public Review):

   The main finding - that the moment-to-moment relationship between excitability and perception is coupled to the body's slower respiratory oscillation - is novel, interesting, and important for advancing our understanding of how the brain-body system works as a whole. The experiment is simple and elegant, and the authors strike the right level of making the most of the data without doing too much and obscuring the main findings. The primary weakness, in my opinion, is the inability to distinguish between the possibility that respiration modulates excitability and the possibility that respiration modulates something boring like signal-to-noise ratio. In terms of conclusions, I thought the authors stuck pretty well to the data. The one place where the conclusions felt a little bold was in terms of the respiration <> alpha <> behavior relationship, where it felt the authors had already made up their minds re: causality. I agree that it probably makes more sense for respiration to influence something about the brain than vice versa, and the background presented in the Intro/Discussion supports this. However, the analysis only tells us that the behavioral performance was modulated by both alpha and respiration (and their interaction, but this is no way causal). Overall, it will be necessary to differentiate the current interpretation from the possibility that breathing and alpha are two unrelated time courses that influence behavior at the same time (and even interact in how they influence behavior, but just not interact with each other), and I do not believe the phase-amplitude coupling analysis is sufficient for this.

   Read the original source
5. 

   eLife

   Nov 29, 2021

   Reviewer #2 (Public Review):

   Kluger and colleagues investigated the influence of respiration on visual sensory perception in a near-threshold task and argue that the detected correlation between respiration phase and detection precision is liked to alpha power, which in turn is modulated by the phase of respiration. The experiments involved detecting a low-contrast visual stimulus to the left or right of a fixation point with contrast settings adjusted via an adaptive staircase approach to reach a desired 60% hit rate, resulting in an observed hit rate of 54%. The main findings are that mutual information between the discrete outcome of hit-or- miss and the continuous contrast variable is significantly increased when respiration phase is considered as well. Furthermore, results show that neuronal alpha oscillation power is modulated in …

   Reviewer #2 (Public Review):

   Kluger and colleagues investigated the influence of respiration on visual sensory perception in a near-threshold task and argue that the detected correlation between respiration phase and detection precision is liked to alpha power, which in turn is modulated by the phase of respiration. The experiments involved detecting a low-contrast visual stimulus to the left or right of a fixation point with contrast settings adjusted via an adaptive staircase approach to reach a desired 60% hit rate, resulting in an observed hit rate of 54%. The main findings are that mutual information between the discrete outcome of hit-or- miss and the continuous contrast variable is significantly increased when respiration phase is considered as well. Furthermore, results show that neuronal alpha oscillation power is modulated in phase with respiration and that perception accuracy is correlated with alpha power. Time resolved correlation analysis aligned on respiration phase shows that this correlation peaks during inspiration around the same phase where the psychometric function for the visual detection task reaches a minimum.
   The experimental design and data analysis seem solid but there are several concerns regarding the novelty of the findings and the interpretation of the results.

   Major concerns:
   The finding that visual perception is modulated by the respiration cycle is not new (see e.g. Flexman et al. 1974 or Zelano et al. 2016).

   There are multiple studies going back decades that show alpha oscillation power to be modulated by breathing (e.g. Stancák et al., 1993, Bing-Canar et al. 2016). Also, as the authors acknowledge, it is well-established that alpha power correlates with neuronal excitability and perception threshold. What seems to be new in this study is the use of a linear mixed effect model to analyze the relationship between alpha power, respiration phase and perception accuracy. However, the results mostly seem to confirm previous findings.

   Magnetoencephalography captures broad band neuronal activity including gamma frequencies. As the authors show (Fig. 4) and other studies have shown, the power of neuronal oscillations across multiple frequency bands is modulated by respiration phase. Gamma and beta oscillations have been implicated in sensory processing as well. Support for the author's hypothesis that the perception threshold modulation with respiration is due to alpha power modulation would be strengthened if they could show that the power of oscillations in other frequency bands are not or only weakly linked to perception accuracy.

   In the discussion the authors speculate that respiration locked modulation of alpha power and associated neuronal excitability could be based on the modulation of blood CO2 levels. Most recent studies of respiratory modulation of brain activity have demonstrated significant differences between nasal and oral breathing, with nasal breathing (through activation of the olfactory bulb) typically resulting in a stronger influence of respiration on neuronal activity and behavioral performance than oral breathing. The authors only tested nasal breathing. If blood CO2 fluctuations are indeed responsible for the observed effect, there should be no difference in outcome between nasal and oral breathing. Comparing the two conditions would thus provide interesting additional information about the possible underlying mechanisms.

   Minor concerns:
   Figures 1, 3 and 4: label fonts are too small on some panels.

   Supplementary figure 3: labels are illegible.

   Read the original source
6. 

   eLife

   Nov 29, 2021

   Reviewer #3 (Public Review):

   The topic is timely, the study is well-designed, and the work has been performed in a highly competent manner. The authors relate three variables: respiration, alpha power and perceptual performance, constituting a link between somatic and neuronal physiology and cognition. A particular strength is the temporal resolution of respiration effects on cognition (continuous analysis of the respiration cycle). Furthermore, results are well contextualized by very comprehensively written introduction and discussion sections (which, nevertheless, could be slightly shortened).

   I have three points of criticism, all meant in a constructive way:

   1. I wonder whether the authors could have gone one step further in the analysis of causal mechanisms, rather than correlations. The analysis of timing (Fig. 4d) and the last …

   Reviewer #3 (Public Review):

   The topic is timely, the study is well-designed, and the work has been performed in a highly competent manner. The authors relate three variables: respiration, alpha power and perceptual performance, constituting a link between somatic and neuronal physiology and cognition. A particular strength is the temporal resolution of respiration effects on cognition (continuous analysis of the respiration cycle). Furthermore, results are well contextualized by very comprehensively written introduction and discussion sections (which, nevertheless, could be slightly shortened).

   I have three points of criticism, all meant in a constructive way:

   1. I wonder whether the authors could have gone one step further in the analysis of causal mechanisms, rather than correlations. The analysis of timing (Fig. 4d) and the last sentence of the abstract suggest that they imagine a causal role of respiratory feedback on cognitive performance, mediated via coordination of brain activity (in the specific case, by increasing excitability in visual areas). This could be made more explicit by appropriate experiments and data analysis:

   1.1. Manipulating the input signal: former studies suggest that nasal respiration is crucial for effects on brain oscillations and/or performance (e.g. Yanovsky et al., 2014; Zelano et al., 2016). Thus, the causal inference could be easily checked by comparing nasal versus oral respiration, without changing gas- and pH-parameters of activity of brainstem centers. Admittedly, this experiment may add significant work to the present data which, by themselves, are already very strong.

   1.2. Temporal relations: The authors show that respiration-induced alpha modulation precedes behavioral modulation (Fig. 4d and related results text). Again, this finding suggests a causal influence of respiration on performance, mediated by alpha suppression (see results, lines 318-320). Could the data be directly tested for causality (e.g. by applying Granger causality, dynamic causal modelling or other methods)? If this is difficult, the question of causality should at least be discussed more explicitly.

   2. At various instances, the authors suggest that respiration-induced changes in pH may be responsible for the changes in cortical excitability which, in turn, affect behavioral performance. In the discussion, they quote respective literature (lines 406-418). I glanced through the quoted papers by Feldman, Chesler, Lee, Dulla and Gourine - as far as I could see none of them suggests that the cyclic process of respiration induces significant cyclic shifts of pH in the brain parenchyma (if at all, this may occur in specialized chemosensory neurons in the brainstem). Moreover, recent real-time measurements by Zhang et al. (Chem. Sci 12:7369-7376) do also not reveal such cyclic changes in the cortex. Finally, translating oscillatory extracellular pH changes (if existent) into changes in inhibitory efficacy would require some time, potentially inducing delays and variance onto the cyclic changes at the network level. I feel that the evidence for the proposed mechanism is not sufficient, notwithstanding that it is a valid hypothesis. Please check and correct the interpretation of the cited literature if necessary.

   3. Finally, some illustrations should be presented in a clearer way for those not familiar with the specifics of MEG analysis. I add some specific suggestions below.

   Read the original source
7. 

   Version published to 10.1101/2021.03.25.436938v1 on bioRxiv

   Mar 25, 2021