Hi Dominik,

You're quite right, many terms are used synonymously, another title that you may have seen is Baroreflex Biofeedback or Blood Pressure Biofeedback. Readers may be limited to discerning the focus of the intervention (and thus the appropriateness of the title) from the reported methods in each study. This is due to a number of factors including (non-exhaustive); the explosion of Biofeedback related studies around the world with differences in publication requirements between research areas, a predominance of proprietary Biofeedback methods for intervention studies (e.g., Heart Math). You will mostly see measures of parasympathetic cardiac activity (e.g., RMSSD) as dependent variables. Again this is due to a number of factors, but perhaps most of all because of the link these measures have with central and peripheral physiology.

The aim of these interventions is to enhance parasympathetic nervous system dominance and alleviate sympathetic over-activation. As you well know, this can be achieved via slow (~0.1Hz) rhythmic breathing which will increase HRV via RSA. Increased HRV (broadly) is indicative of, and causes increased Baroreflex activity (blood pressure variability) which is widely considered the main therapeutic mechanism underlying these approaches. Herein lies the problem because the following can be termed in many ways; prescribing or training respiratory patterns or frequencies to increase RSA levels and HRV amounts to exercise blood pressure variability or Baroreflexive mechanisms.

From what I have seen Coherence and related terms are attributed to proprietary methods whereas HRV and RSA are closer to clinical terms, although they may all be very similar in delivery. I have found Baroreflex Biofeedback and similar to be more common in clinical publications. Respiratory biofeedback could vary from simple respiratory rate feedback (i.e., during tasks/induced stress Vs relaxation), or also include HRV and/or blood pressure variability measures.

Delivery techniques vary between self-initiated and prescribed methods which may also play a part in the terms used. For example, some interventions increase RSA by asking the participant closely following HR changes with their respiration, whereas others will require participants to simply follow a breathing pacer. In both cases HRV may be the dependent variable and respiratory rate the independent variable, however the method to achieve the desired physiological effects differ. These differences may go some way to distinguishing Biofeedback from Training, however, as far as I know there has yet to be a consensus on the correct way to distinguish these. Arguably all Biofeedback techniques require training, or some form of bespoke education and learning.

Regarding optimal frequencies, it appears that maximal HRV measures (especially those expressed in Time Domain) may occur at around four breaths per minute, as opposed to six breaths per minute (0.1Hz). This does not necessarily suggest that Baroreflex activity is optimally exercised at four, which is perhaps indicated by drops in RSA as you've noted. Other issues include the ratio of inhale to exhale and breath volumes, both of which have have received little attention.

I hope this information helps, apologies for any poor spelling and grammar. Feel free to ask other questions. For further information I suggest you see 'How and why does HRV biofeedback work' Lehrer & Gevirtz (2014) and the 'Neurovisceral Integration Model' by Thayer et al., (2000).

Regards,

Mike

Here's a document summarizing HeartMath's Coherence algorithm.

https://www.heartmath.org/assets/uploads/2016/04/coherent-heart-integral-review-2009.pdf

See page 23, Heart Rhythm Coherence Ratio Calculation.

To understand the different mathematical descriptors of heart rate variability (HRV) associated with breathing rate, read these two references:

Grossman, P., & Taylor, E. W. (2007). Toward understanding respiratory sinus arrhythmia: relations to cardiac vagal tone, evolution and biobehavioral functions. Biological Psychology, 74(2), 263–285.

Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. (1996). Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation, 93(5), 1043–1065.

Note that any muscle (EMG) artifact or premature ventricular contraction (PVC) that occurs while HRV is being scored will destroy the HRV calculation unless the interbeat interval data associated with the PVC or EMG artifact are cleaned by interpolation.

Also, note that LF and RSA are influenced in complex manners by both branches of the ANS. The 0.15 Hz dividing line between LF and HF is certainly not an absolute cut-off. Thus, some sympathetic effects may also impact the HF portion of the HRV spectrum.

Thank you for the helpful answers!

- I wonder if HRV-Biofeedback is technically speaking really Biofeedback. Once the "resonant frequency" is found, its merely breathing with a pacer, isn't it? In other words, if you do this kind of training to increase HRV (e.g. with live played music, like i am trying right now in a study) without searching for the resonant frequency or in just a single session, you (i.e. I) would have to call it training, or is it still within the scope of HRV-Biofeedback?

- The resonant frequency is said to be the individual breathing frequency between 4,5 and 6,5 breaths/minute, that produces the highest HRV. But research indicates, that the highest HRV is generatet around 4 breaths/minute? But then 4 is presumably not the healthiest frequency, that is, not as healthy as the resonant frequency at around 6 breaths/minute.

- Is the frequency analysis in HRV-Biofeedback thus more important, than in other biofeedback methods? What we "want" is the peak at about 0,1 Hz. But what exactly is good and healthy about that large peak (in the rather sympathetic LF-frequency band) and the underlying coherence with the breathing opposed to a heterogeneous frequency spectrum? 

Thank you!

Hi Dominik,

1. I agree, finding a resonant frequency is not in itself training. I'd argue that it is soft-core biofeedback in that the feedback is provided by a practitioner, perhaps at a later time/date. For lack of a better term, 'hard-core' biofeedback would be a more exploratory process where the individual receives feedback in real-time to which they must develop techniques to remain within certain parameters or thresholds. In my work with elite sports personnel, individuals are required to accurately maintain their resonant frequency for five minutes without a pacer. This is the end-intervention aim preceded by easier goals & plenty of feedback and suggestions.

2. The highest amounts of HRV at four breaths/minute are normally demonstrated in time domain statistics (e.g., SDNN, RMSSD). When using FFT frequency analysis (not without significant short-comings) the total power and LF band should (generally speaking) record higher values at six as opposed to four breaths/minute. One reason may be that mean R-R intervals are longer at four breaths per minute during exhalation, essentially providing less data-points for FFT and larger SD of intervals. Anecdotally, I use four breaths/minute to demonstrate the physiological control individuals can achieve via breathing (i.e., feeling very sleepy), but have found all resonant frequencies to be between 5 and 6.7 breaths/minute.

3. Contemporary HRV understanding has moved from LF = sympathetic & HF = parasympathetic, posing a problem for previous analyses that have relied on the ratio LF:HF. The LF band is influenced by both wings of the ANS to differing amounts, dependent on many individual and situational factors. There are some studies suggesting that the LF band is a measure of autonomic cardiac baroreflex outflows, rather than sympathetic cardiac activity (e.g., Goldstein et al., 2011). During RF breathing, a peak at the corresponding frequency (e.g., 6 breaths/minute = 0.1Hz) is indicative of the power of RSA, or the amount of HRV determined by respiration. This is the intended (acute) effect and I have seen particularly fit individuals achieve a peak representing ~60% of total power, the inference is that the Baroreflex is being stimulated. Outside of paced breathing, healthy HRV is chaotic (in FFT representations) although there may be distinguished peaks in each frequency band.

I hope that helps. Thanks,

Mike

Thank you, Mike. 

So the resonant frequency is the one with the highest peak at around 0,1Hz and not the frequency with the highest outcome of RMSSD? I still haven't entirely understood what the immediate health benefits of the resonant frequency - as opposed to 6 breaths/minute for everyone - are. Is it the coherence of respiration, heart rate and baroreflex or pulse transit time resp.?

In our case, the breathing at around 0,1Hz is facilitated by a Music Therapist via Monochord (Inhalation) and Chanting (Exhalation), which seems to be a easy to understand and well received way of pacing. Though the Music Therapist sees and works with the visual feedback, I am not sure if it is accurate to call our intervention (single session) HRV Biofeedback, since it is more like a HRV/RSA Training?

Hi Dominik,

Our experience with RSA and controlled breathing over a few hundreds of subjects (from 11 to 85 years) conducted to the following observations:

- Most of people are able to follow a 6 cycles/min breathing pace

- Some of them (~15%) are not able or they don't feel comfortable with this pace, so they ask for a faster pace   

- RSA at with a controlled respiration at 6cycles/min is almost always visible on the spectrum as a peak at 0.1Hz (excepting for subjects having various arrhythmias). Not entering here into LF/HF debate, especially with RSA. 

We believe that there is a resonant breathing frequency for each body (around 6 cycles/min) and we believe that there is a relationship between that and the heart rate value. We discovered that during a meditation experiment for ~2 hours, as shown in the attached document. The heart started suddenly to oscillate around a heart rate of 57bpm, then 10 times the spontaneous respiration which was at 5.7 cycles/min.

As this particular physiological state is associated with an exceptional wellness state.

So, in my view, it is better to have a close look at the frequency at which the peak is located, not at its amplitude level. Relationship to the average heart rate can be interesting as well. Regarding the amplitude of the RSA, is better to calculate it's RMS value in time domain but I don't think it have a great importance on choosing the right breathing.  

If there are any questions related to attached experiment and RSA, let me know.  .                         

Vasile, thanks for that paper and your website, I'm downloading your app. Great contribution to the HRV field. Have you considered interfacing in realtime to R-R monitors such as the Polar?

Dominik, Here's what I've noticed with the HeartMath emWave technology. You can utilize their built in (visual) breath pacer initially, but my impression is that any 'paced' breathing patterns are NOT really biofeedback. They might help to ingrain future resilience to stress situations, but this is not as efficacious as HeartMath's other suggestions, which are:

To use the idea of holding an intention of gratitude or appreciation, felt in the heart area during your meditation time with the device. In this mode, you can actually turn off the breath pacer visual and audio and just go with the sensations you are noticing in your chest / heart area. You then gently take note of the LED color (red, blue or green = highest coherence.) This is what actually builds resilience over time. It is not mechanical like the breath pacer, but neuroplastically creates new modes of perception / interoception.

If you saw that pdf link on my previous post, their FFT algorithm adapts to the natural breathing pace, rewarding for purity of the coherence sine wave signal. It is not dependent on the actual frequency or amplitude.

https://www.google.com/search?q=interoception

Thank you William. For now, CODESNA_HRV is intended for researchers wanting to have a complementary tool for HRV assessment that provides real times ANS dynamics and new markers, compared to existing tools. Later on, we'll check the opportunity to interface with commercial monitors in real time.

Hi Dominic

your question is quite old but I just found it and want to answer. I am a psychologist and a biofeedback-trainer doing biofeedback training all day long. I don't like all the terms, what I use is (german) Kohärenztraining (engl.: training of coherence). HRV says only something about "are there any changes in the pulse frequency", RSA says something about the form of the changes of the pulse frequency and my term "Kohärenz" includes other levles of sns and psns output e.g. pulse amplitude.

As a biofeedback-trainer and not a researcher I see the form of the curves online while the patients are producing it , pause the mesurement and talk with them about details of the curves. So the algorithms how to calculate HRV or RSA are not so important for me.

@Wiliam: thank you for your contributions, I agree very strong with you: any pacing is just the beginning to lead the pations in a quick way towards the aim-state. Deeper balance is just possible when the people feel this state and the rhythm in themselfes.

@Vasile: thank you for the papers!

KR (LG), Wolfhard

There is similarity in HeartMath approach with Relaxation Response (Dr Herbert Benson's findings). Adding "feeling" (from the past positive experience) could transform. I use Emwave Pro and I do notice "10" seconds though there are individuals on the either side (Depression-below 10 seconds, sportspersons above 10). Also, relaxation (similar to what HeartMath says) is different compared to positive emotion based (e.g. prayer) breathing which would provide a more complete spectrum (i.e. higher peak). In short, feeling+breathing=higher HRV. (not just breathing nor just feeling alone). Does this understanding make sense?

Hi Gunjan,

what you said is exactly my experience. Breathing allone is not enough, feeling allone is not enoug. Both must fit together in a very special way. I can specify "how to feel". The most important part seems to be the quality of beeing able to "stopp all activity" while exhaling. Calm down, "let it be" while exhaling, and then very light activity, beeing happy while inhaling.

Late to the party, but I think you raise an important question and I'm honestly surprised that the term HRV biofeedback continues to be used so loosely. In my mind, for biofeedback to be effective, it requires rapid feedback of the biological signal. This is harder for HRV than EEG, HR, skin conductance, etc., because HRV traditionally requires a longer sample time in order to measure the variability. The work by Boris Bornemann is interesting because it is an attempt to reduce this time and provide a measure of HRV that can be more immediately fed-back. I think this is the only sort of approach that should be termed biofeedback, but it seems the record of published work in this area is against me. Article Helping from the heart: Voluntary upregulation of heart rate...