Hi  Ali,

Good question! In short, LF/HF results are meaningless in that case, please find below the explanation.       

Slow breathing case like 0.1 Hz (~6 cycles/min) is a particular case when ANS is almost completely modulated by the respiration. Mathematically speaking, in that case almost all the power of the HRV is contained around the respiratory frequency (0.1Hz) and the spectral analysis will show a peak at 0.1Hz, like for sine wave type signal. This is normal since RR signal here looks like a sine wave, deterministic signal. Regarding the LF/HF analysis performed in that case, you have noticed that LF is very high and HF very low; and this is normal since almost all the signal power falls into LF band, all the power being around 0.1HZ. Therefore, analysing LF/HF (normalized or not) is just meaningless here since it cannot bring you any useful meaning, all you can see in the spectrum is the peak at 0.1Hz corresponding to the respiratory frequency.

Therefore, time domain analysis is much more useful, especially non linear analysis as it will provide the ANS balance of each individual. The amplitude of HRV (RMSSD, SDNN) is much higher at a resp of 0.1Hz (6 cycles/min) than in the baseline case or even compared to 0.2 Hz breathing frequency; and this high amplitude is often misinterpreted as a level of vagal activity. In reality, vagal activity is activated only during exhalation (PNS) since vagus nerve is released during inhalation. Having a high amplitude of HRV (SDNN and RMSSD) is good physiologically (and emotionally) speaking, but having a well balanced sympatho-vagal response is also important in long term since the amplitude of HRV decreases with aging.             

As of today, CODESNA_HRV is the only scientific tool able to analyse HRV in such way and to provide the true ANS dynamics and balance, especially in slow breathing case, but also whatever wanted protocol.          

I have attached a short document analysing slow breathing use case by using the FFT and also using our own markers, that illustrate some of comments I mentioned above. 

Hope this helps!

Hi Dr. Gholamrezaei,

Dr. Zoicas provides a good explanation. I would just add that what you found is not uncommon (e.g., Eddie & Kim et al., 2014 attached). You also may find the following discussion thread useful, 

https://www.researchgate.net/post/Is_it_possible_to_calculate_the_coherence_of_Respiration_Rate_and_Heart_Rate_with_only_the_Respiration_Rate_and_the_frequency_spectrum_at_hand/1

I am pleased to see Dr. Zoicas may have a better way to analyze ANS balance. There is an ever growing literature suggesting we should stop reporting the LF/HF ratio as a measure of sympatho/vagal balance, e.g., http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3576706/

One way we've been analyzing shifts in ANS balance is using the D metric (See 'D Statistic Article' attached). The MatLab syntax to calculate this statistic is freely available on the web,

http://aimdyn.com/W1index/

Dear Vasile,

Thanks for your response. You have claimed that the CODESNA is the only system that provides real-time PNS and SNS activity. Is there any physiologic evidence (not just mathematical algorithms!) approving this claim? for example any physiological data by ANS nodal blockade? or from animal experimental studies by ANS denervation? As I know, NeuroScope also has the same claim providing real-time vagal activity. However, in the case of NeuroScope, atropine is used to provide a scale for vagal activity (http://www.ncbi.nlm.nih.gov/pubmed/9950879).

Dear, Ali!

Interpretation of the autonomic regulation by HRV during spontaneous breathing less than 10 times / min, excluding tidal volume, is difficult. At the same time, the use of the test in the form of 6 breaths per minute, together with spontaneous breathing allows the interpretation for baroreflex sensitivity. To assess HRV during breathing less than 10 times / minute, we have developed a nomogram. You can also use spreadsheet estimates for 6 breaths per minute. In this case, the physiology is difficult.

Dear Ali,

Indeed, several researchers already use CODESNA_HRV as the main tool for the ANS assessment. A first evaluation and validation (in french only for now) was made by Dr. Mourot (http://www.codesna.com/en/scientific-comitee/) and a couple of studies are ongoing, including elite athletes and hospital patients. Official outcomes and publications will go out in 2016. Meanwhile, you (and any researcher here) are very welcome to evaluate the validity of CODESNA results by yourself. The SW is very easy to download, to install and to use and I guess you have several RR files  that you might want to evaluate in a different way than LF/HF. Please check it out here: http://www.codesna.com/en/products/codesna_hrv/  User guide and RR interval examples files are provided with the SW.

Dear Dr. Eddie, thank you for the insight and links.

Dear David,

Thanks for the comment and introducing the articles. I would like to add the review by Goldstein et al. on evidence against LF and LF:HF being valid markers of cardiac sympathetic tone. http://www.ncbi.nlm.nih.gov/pubmed/21890520

Vasile, you may would like to have a look at this review in which some markers and methods that have been used for the validity check of HRV analysis are mentioned, e.g. cardiac noradrenaline spillover, sympathetic neuroimaging, ganglionic bloockade, pharmacologic denervation, known group validity, etc. This is indeed a big claim that CODENSA can validly mmeasure ANS activity, but its validity must be evaluated against such markers. Otherwise, its just an unapproved assumption based on theories !

Thanks Ali, that's a very useful article.

We can use The Model IPFM.And Partly to help understand us Autonomic nervous system or autonomic nervous system, system dynamics that regulate smooth muscle function, muscle, heart and endocrine related. In other words, this part is of the peripheral nervous system (PNS) 

In IPFM model there is an input signal which is fed to an integrator and when the output of the integrator reaches to a threshold then we have a pulse. This pulse simulates the R-wave of the ECG signal. The source of ECG is sinoatrial (SA) node. SA node is affected by many factors. The power spectrum of HRV signal is divided into three main bands i.e. VLF, LF and HF. The information in these three bands shows how the Autonomic Nervous System (ANS) affects on heart activity. The ratio LF/HF is related to sympathovagal balance. The proposed model generates both normal HRV and HRV of some abnormalities with regard to ANS. As we used random signal in the input of the IPFM the output of the model has random variations and the output never be as a periodic signal. Our model is very similar to real HRV than a normal IPFM.

See here how we interpreted cardiovascular responses to slow deep breathing

https://doi.org/10.1111/psyp.13447

Vasile Zoicas I dont think, one can claim that slow breathing does not activate a vagal tone. Paul Lehrer, McCraty R has written extensively on this topic. Can you share papers that refute their claims?

Debanjan Borthakur tnx for your comment. please see our paper here

Article Influence of inspiratory threshold load on cardiovascular re...

When we say 'activate' 'vagal tone' we must be very objective. what do we mean by 'activate' and by 'vagal tone'?

slow breathing certainly modulate cardiac vagal activity and we can see this by increased variability of HR at the frequency of breathing, and pharmacological blockade studies support that this is mainly mediated by cardiac vagal modulation. However, pharmacological vagal blockade is not selective regarding vagal phasic or tonic activity see this

Article Vagal Mediation of Low Frequency Heart Rate Variability Duri...

Almost all studies so far have shown that during slow breathing and in a single session there is NO significant alteration in average HR which doesn't support alteration in cardiac vagal tone, at least not in short term. This may be different if slow breathing is practiced for a long term, and mechanisms may also be different, particularly emotional modulation may also be involved.

Also see this for vagal phasic vs tonic activity

Article Toward understanding respiratory sinus arrhythmia: Relations...

Ali Gholamrezaei Slow breathing increases SDNN (Paul M. Lehrer, et al. 2014), Increases RMSSD ( Fuchs et al, 2018). Even with my own study, I found an increase of HRV in 5 minutes of slow breathing.

Long term practice of slow breathing also increases sdnn ( Hideaki Hasuo et al, 2020). Bin Yu, 2018 also used SDNN as a measure of HRV for their app.

Debanjan Borthakur

1- SDNN is not an index of pure cardiac vagal activity, it can also include sympathetic influence since it is indicating an overall variation in HR, not beat to beat variation which is fast and almost entirely vagal

2- RMSSD is an index of cardiac vagal modulation not tone, since it represents beat to beat variation, although UNDER RESTING STATE there is high correlation between RMSSD and average HR.

3- RMSSD under SLOW BREATHING is not correlated with HR, since RMSSD under slow breathing largely is influenced by cardiac vagal phasic modulation at the frequency of breathing but average HR is not altered.

In all of our experiments with slow breathing (>10 experiments), slow breathing alters all index of HRV, but not the average HR which indicates that in short term slow breathing increases MODULATION of cardiac vagal activity (through baroreflex and partially via pulmonary vagal afferents, see our paper in previous comment) but not cardiac vagal tone. If average HR also decreases then it can indicate alteration in cardiac vagal tone.

Please read this for distinguishing btw cardiac vagal phasic and tonic activity

Article Toward understanding respiratory sinus arrhythmia: Relations...

Also see this amazing animal mechanistic study

Article Brainstem sources of cardiac vagal tone and respiratory sinu...

=> cardiac vagal modulation ~= cardiac vagal tone

=> any intervention that increases cardiac vagal modulation increases HRV but does not necessarily increase cardiac vagal tone; why because HRV is not vagal tone!

=> breathing exercises although increases cardiac vagal modulation (increase in RSA) does not necessarily increase cardiac vagal tone

See here some of our papers:

Article Can Slow Deep Breathing Reduce Pain? An Experimental Study E...

Article Respiratory Hypoalgesia? The Effect of Slow Deep Breathing o...

https://pubmed.ncbi.nlm.nih.gov/31361032/

https://pubmed.ncbi.nlm.nih.gov/25156003/

Unfortunately the literature on HRV, especially psychophysiology literature, can be miss-leading. To interpret HRV, we always need to check original papers that has compared HRV indexes against a valid measure of ANS (and under different context) or animal studies using central/peripheral neural blockades to investigate the exact mechanisms. Moreover, interpreting HRV largely depends on whether breathing behavior is altered or not.

Hope these can help

Thanks Ali Gholamrezaei . How do you differetiate between vagal tone and vagal modulation? How to quantify vagal tone then? The wikipedia seems to disagree with your view as you said, '' RMSSD is an index of cardiac vagal modulation not tone ". I will quote,

'There are several methods of estimating vagal tone other than measuring RSA, including: Indexes of beat-to-beat variability such as RMSSD reported by The Task Force of the European Society of Cardiology and Heart Rhythm Society.

Frequency analysis of heart rate in the range 0.15–0.4 Hz has been reported to quantify vagal tone.

Debanjan Borthakur As a scientist, I do not consider Wikipedia as a valid source of evidence when it comes to understanding the physiology. You need to always READ THE ORIGINAL PAPERS, as you see the Wikipedia has wrongly interpreted the 'Task Force of The European Society of Cardiology' guideline. In the guideline, it is clearly mentioned that "It should be remembered that the components of HRV provide measurements of the degree of autonomic modulations rather than of the level of autonomic tone [28] and averages of modulations do not represent an averaged level of tone." Also please read other references I mentioned above.

Article Toward understanding respiratory sinus arrhythmia: Relations...

Article Brainstem sources of cardiac vagal tone and respiratory sinu...

Same problem with spectral measures of HRV, they are even more influenced by alteration in respiration frequency, as we and others have shown in multiple experiments. Please read the original papers, not the Wikipedia.

Another thing I would recommend is to read basic human physiology textbook first if you do not have such background. This will help you a lot in understanding the literature and interpreting the physiological data. You can start with this one:

https://www.us.elsevierhealth.com/guyton-and-hall-textbook-of-medical-physiology-9781455770052.html

If you read this paper, ''Article Toward understanding respiratory sinus arrhythmia: Relations...

In slow breathing, RSA amplitude maximizes, so we have a strong argument in favor of resonance breathing to be para sympathetically arousing. Let me know if you disagree.

Ali Gholamrezaei I think there is a misunderstaning. I agree with you RSA might not represent vagal tone ( although most of the researchers used it) also lnHF is considered as marker of vagal tone. Now, you asked me to read one paper and their conclusion was that, ' HRV that occur during slow paced breathing by showing that changes in low frequency power under these conditions are almost entirely vagally-mediated. Slow paced breathing is an effective tool for cardiac vagal activation. ' So, it's obvious that the increase in LF or SDNN or RMSSD is actually pointing towards an increased vagal tone, provided the papers you cited are correct. I hope , we both agree with this.

Debanjan Borthakur

1- Let's read these papers more carefully. Laborde et al. has mentioned in the introduction that "we refer to parasympathetic activity as vagal tone" that means that they are mixing all types of PNS activities together and call it vagal tone. So when reading this review keep this in mind. Also, they have mentioned that ""in this paper we refer to cardiac vagal tone as assessed by HRV measurement" well we cannot define a concept by a surrogate measure and use it as a method of validation. We need to always look at the gold standards; neural blockade studies and pharmacological blockade studies.

2- Grossman et al indeed have mentioned that "RSA has been shown to often provide a reasonable reflection of cardiac vagal tone when the above-mentioned complexities are considered", but also have listed 6 issues above of this that "can lead to misinterpretation of cardiovascular autonomic mechanisms" and one is respiratory parameters, please also see others. These are mostly overlooked by many researchers in psychophysiology.

3- HF, LF, or SDNN or RMSSD interpretation as a measure of cardiac vagal modulation requires paying attention to the issues raised by Grossman et al. When respiratory frequency changes from HF to LF, HF no longer represents the respiratory modulation of cardiac vagal activity, but LF does, as was shown by pharmacological blockade study. This doesn't mean that under resting state and with respiratory frequency being in HF band, LF still can represent vagal activity, this also has been shown by pharmacological blockade studies.

4- To interpret HRV, you need to first look at the physiology of HR modulation by ANS at the level of the heart and at cell level, as well as interaction between SNS and PNS at central and peripheral levels. As mentioned by Laborde et al "ease of access [and of measuring HR] should not obscure the difficulty of interpretation of HRV findings that can be easily misconstrued". We have too many studies using HRV to measure ANS activity but too little studies on the validity of such measures.

Finally, that many researchers repeatedly mentioned something does not mean that is correct. This is the problem with HRV interpretation in psychophysiology literature, unfortunately, and we continue repeating that without asking ourselves based on what evidence and based on what valid or gold-standard measure? If you look at cardiovascular physiology literature you usually don't see such misinterpretations. Same with 'resonant breathing', nice phrase but no strong evidence.

Ali Gholamrezaei Thanks for your response. I almost entirely agree with you, and I really appreciate you for delving into this discussion. InArticle Heart rate variability biofeedback: How and why does it work?

Ali Gholamrezaei I read few papers that contradicts your views. I am not an expert in this field, just trying to gain some insight. Shaffer (Article A healthy heart is not a metronome: An integrative review of...

Please check out this use case and you will see the spectral analysis is useless in that case.

Research Case study on HRV – Real time monitoring of meditation shows...

Debanjan Borthakur I'm not saying the slow breathing does not activate "vagal tone" although the definition of this is still confusing. We did a lot of experiments on slow breathing and event though vagus nerve should be dominant and stimulated in that case, you will always get the LF higher than HF. That's why I'm saying the LF/HF analysis are meaningless in that case. Please have a look on what happens in real time with ANS on the use case above using our method and you will see that clearly. stressed people cannot activate vagus nerve fast even with slow breathing.

Hi Ali,

I have found the following review helpful: Article The physiological effects of slow breathing in the healthy human

Best,

Florian

Yes, Florian is correct. I too found that the article "The Physiological effects of slow breathing..." so helpful to get an basic idea. Me too is planning to record HRV during slow breathing during my PhD data collection.

Best Regards,

Udani

Hi

Thanks for all the suggestions

To give you an update here I list our publications where we give an interpretation of HRV in response to controlled breathing and mechanisms underlying effect of controlled breathing on pain

https://pubmed.ncbi.nlm.nih.gov/33111377/

https://pubmed.ncbi.nlm.nih.gov/31978501/

https://pubmed.ncbi.nlm.nih.gov/31698133/

https://pubmed.ncbi.nlm.nih.gov/31361032/

I have limited knowledge and virtually no research experience with HRV. However, I have used it personally for 20 years, mostly out of curiosity.

The "paced breathing" - 5 seconds inhale, 5 seconds exhale - is more than I normally breathe. I have noted one interesting phenomenon: depth of inhalation has an effect. I can change results by controlling it. Just an observation.

Hi Ray Barnum Thank you for sharing your experience. Indeed that is what we found in our experiment here Article Influence of inspiratory threshold load on cardiovascular re...

and here

Article Psychophysiological responses to various slow, deep breathin...

to take a deep breath, we need to increase the air flow and in order to do that we need to increase the negative pressure in chest. That results in larger initial drop in blood pressure, increased blood return to heart, and subsequent responses via the baroreflex. The pulmonary stretch receptors also play a role but seems not as much as the baroreflex.

The best technique is Respiratory sinus arrhythmia (RSA), it is heart rate variability in synchrony with respiration, by which the R-R interval on an ECG is shortened during inspiration and prolonged during expiration.

My training in physiology is limited, but this is an area where the more I learn, the less I understand. Yours are the sorts of questions that need to be examined, but unfortunately seem to be glossed over by so many.

I found Sevoz-Couche & Laborde's recent (2022) article is an interesting read if you haven't seen it already.

"Essentially by slowing breathing rate, we are coupling our respiratory and cardiac rhythms, and this phenomenon is known as the resonant frequency. Accordingly, when the breathing rate is 0.1 Hz, all natural and applied cardiac oscillations appear at equal frequencies, and the amplitude of applied cardiac oscillations is maximal"

"Kromenacker and colleagues (2018) showed that HRV power across the low-frequency range was found to be nearly eliminated by the parasympathetic blockade during SPB, while similar spectral power during sympathetic blockade and placebo was statistically indistinguishable, reinforcing the fact that RSA is vagally-mediated even during SPB."

But, assuming one can increase RSA amplitude through this practice (whether at a general 6BPM or at an individualized "resonance" frequency) the question remains: why does increasing RSA (acutely/sporadically/chronically) have any impact on physical health/mental state. Sevoz-Couche & Laborde touch on this, but it still seems to be a rather weak area.

And, just to throw this into the mix: in addition to the effects of respiration rate on HRV measures, should some sort of correction be used for the effects of differing HR?

de Geus, E. J. C., Gianaros, P. J., Brindle, R. C., Jennings, J. R., & Berntson, G. G. (2019). Should heart rate variability be “corrected” for heart rate? Biological, quantitative, and interpretive considerations. 56(2), e13287. https://doi.org/10.1111/psyp.13287

Sevoz-Couche, C., & Laborde, S. (2022, Feb 12). Heart rate variability and slow-paced breathing: when coherence meets resonance. Neuroscience and Biobehavioral Reviews, 135, 104576. https://doi.org/10.1016/j.neubiorev.2022.104576

Grant Benham Thanks for your reply. Actually I put this question on 2015, that was the beginning of my PhD :) then we did multiple experimental studies and of course reviewed other literature and discussed this in our papers, for example

https://pubmed.ncbi.nlm.nih.gov/33111377/

https://pubmed.ncbi.nlm.nih.gov/31978501/

https://pubmed.ncbi.nlm.nih.gov/31698133/

https://pubmed.ncbi.nlm.nih.gov/31361032/

whether the increased RSA is a parallel outcome or a mediating mechanism for the effect of slow breathing is still an assumption and there is no direct evidence supporting that. there is even not enough evidence that such an effect is long lasting. I think there is too much emphasis on RSA and frequency, maybe because it is visible in short term, and we are missing other important aspects of this breathing exercise like emotional and cognitive components. I avoid to use the term resonant frequency breathing because I am not convinced that the effects/effectiveness is due to the so called 'resonant' effect. it is a very catchy term though.

My opinion on paced respiration is just opposite to the idea that we need to control respiration. Both rhythms are so flexible and so tightly coupled, that fixing any of them alters both. Of course there are many meaningful results under such conditions, but we should be aware that the system is no longer intact.

Teodor Buchner interesting and worth empirically testing

My friend, a physiologist, observed that people respond in a typical way, trained since early childhood, and when you tell them that you monitor their respiration, they change respiratory pattern and start voluntary deep breathing. So one of advises in order to observe spontaneous respiration would be not to inform the subject directly that his/her respiration is just being monitored.