Response to: Resonance Frequency Assessment: The Impact and Implications of Inaccurate Assessment in the Clinical Use of Heart Rate Variability Biofeedback

Abstract

Although the paper by Kohlenberg (2021) raises some cogent questions about the necessity for accurately assessing resonance frequency (RF), the argument is an oversimplification and, we believe, underestimates its importance. Several issues are involved: (a) does breathing at exact RF produce greater amplitudes of heart rate variability (HRV) than simply breathing at six breaths per minute, (b) does breathing at exact RF have any clinical advantage, and (c) can RF be accurately determined using currently accepted procedures?The answer to this question is yes. Transfer functions we have published previously (Vaschillo et al., 2002; Vaschillo et al., 2004) show a gradual increase in HRV amplitude as breathing approaches RF, with one frequency producing the highest HRV amplitude. The same is true for baroreflex gain. This is the RF.There are some statistical data that bear on this, but not enough. One small study of treatment for borderline hypertension (Lin et al., 2012) showed nonsignificant but observably greater decreases in blood pressure for the exact RF procedure compared with six-breaths-per-minute breathing. With greater power, if reliable, a finding of this size could have some clinical implications. Another study (Steffen et al., 2017) compared the therapeutic effects for RF to RF plus one breath per minute for mood and blood pressure. Results indicated that:For blood pressure, both the RF and RF+1 group showed reduced blood pressure during breathing training, but the RF group showed the smallest increase in BP in a stress trial (see Figure 1). That is, the results suggest that “RF breathing buffered the stress response” to the stress challenge (Steffen et al., 2017, Impact of Experimental Group on Blood Pressure and Impact of LF/HF Ratio on SBP and DBP section, para. 2). We do not know if the difference could be sufficient to prevent major medical problems such as stroke or heart attack in someone who is hypertensive, but at the cost of only a little training to find exact RF, why not use a method that lowers blood pressure even a little more? Of course, we don't yet know if using exact RF would make any difference for treatment of stress, anxiety, depression, asthma, pain, etc., although it appears possible that it would.The paper by Dr. Kohlenberg makes no reference to the procedures used, or the amount of training that assessors had. What criteria did assessors use? It is true that some criteria we have previously advised (Lehrer et al., 2013) may be less important than others. For example, phase relationship between HRV and breathing tends to break down in older people (Lehrer, 2020), perhaps because of changing delays in cardiac response to baroreflex or respiratory stimulation. We have found that the frequency producing the greatest HRV tends to show a greater phase delay in older people, with respiration starting a few seconds after a heart rate fluctuation occurs. So, phase relationship may not work all the time as a measure of RF and may be at variance with other indices. Also, particularly when measured in the first session, the novelty of doing slow paced breathing may produce emotional changes over time in an individual, such that as a session progresses, if the person relaxes, HRV may increase independently of respiration rate, so it becomes difficult to differentiate effects of changing emotional level from resonance effects. There tends to be greater stationarity of response in more experienced trainees. In one study (Vaschillo et al., 2006), we found stable and significant negative correlations between RF and height—which we have interpreted as an index of blood volume—only after four sessions of training, suggesting that assessment of RF in the first few sessions was unreliable. This is why multiple sessions of testing is essential for accurate RF determination. The fact that this is often not done in the clinical situation just reflects clinicians' reluctance to spend the time at it and willingness to accept smaller resonance effects. In the first session, differences among breathing frequencies may produce unreliable effects, such that a repeated trial of breathing at a frequency initially found to produce the highest HRV amplitude may produce smaller amplitude later.Then, there is an additional consideration. Most procedures to find RF have the trainee breathe at frequencies separated by 0.5 breaths per minute (4.5 breaths per minute, 5, 5.5, 6, 6.5, etc.). Rarely does exact RF fall at exactly one of these frequencies. Where it occurs between two of them, it would not be unreasonable to find that determination of RF would change by 0.5 breaths per minute from time to time, because this is within the error of measurement. We have published this observation (Vaschillo et al., 2006). Other methods have been proposed for more accurate determination of RF, including one that tests at very small increments of respiration rate (Fisher, 2019), but these require further validation.Ultimately, the most accurate way to get people to breathe at close to their exact RF is to use the final procedure recommended in our protocol: to use biofeedback (Lehrer et al., 2013). After determining the approximate RF, we have people follow the HRV tracing and adjust their breathing to produce the greatest HRV amplitude. This should guide the individual to breathe at the exact RF.Therefore, in our practices, we do assess RF in more than one session and follow this by biofeedback. We cannot unequivocally say that this is a superior method to breathing at a predetermined RF. However, if the clinician does choose the latter method, we suggest that they not use six breaths per minute as the standard. We have found that the average RF actually is 5.5 breaths per minute (about 11 seconds), with slower average RFs among individuals with greater blood volume (taller people and men) than those with less (shorter people and women), with ranges from 4.5 to 6.5 breaths per minute (Vaschillo et al., 2006). Other populations, such as children, may struggle with breath rates as slow as 6 or 5.5 breaths per minute. Furthermore, some folks are more comfortable with a 50% inhale, 50% exhale and thus may need feedback to find the right pace.While the Kohlenberg article raises a number of salient points regarding RF in HRVB, pending further research, it would seem wise for the biofeedback practitioner to follow the conventional guidelines (Lehrer et al., 2013).