Heart Rate Variability Biofeedback for Tinnitus: Preliminary Findings from Multiple Case Studies

Biofeedback Volume 43, Issue 3, pp. 142 148 DOI: 10.5298/1081-5937-43.3.04 FEATURE ARTICLE ÓAssociation for Applied Psychophysiology & Biofeedback www.aapb.org Heart Rate Variability Biofeedback for Tinnitus: Preliminary Findings from Multiple Case Studies Andrea Meckley Kutyana, MA, BCN, QEEGT Bio-Optimized, Asheville, NC Keywords: tinnitus, heart rate variability, biofeedback 142 According to the American Tinnitus Association, up to 30 million people suffer from tinnitus and, of those, 12.2 million experience tinnitus severe enough to warrant medical attention. Tinnitus is believed to result from an abnormal auditory perception reflecting dysregulation of the central (CNS) and autonomic nervous system (ANS). However, regulating the ANS has received very little research attention despite the fact that stress is correlated with exacerbation of symptoms and distress. It is believed that when the autonomic nervous system is calm, the presence of severe tinnitus will be less noticeable and individuals can shift their experience from one of severe debilitation to one of acceptance and peace. Three case studies are presented as an initial investigation into the impact heart rate variability biofeedback may have on the subjective perception of tinnitus and the accompanying distress. Further research is needed, but heart rate variability biofeedback may prove to be an effective adjunct intervention for tinnitus. Introduction Tinnitus is generally referred to as a hissing, sizzling, buzzing, or ringing in the ears. More formally it is defined as the conscious perception of a sound that occurs without an external source (Tunkel et al., 2014). Tinnitus is a fairly common occurrence with 30 million U.S. adults reporting they suffer from the condition (Kochkin, Tyler, & Born, 2011). While the majority of individuals are able to cope with, or essentially ignore, tinnitus, many are emotionally impacted and a small percentage are found to be completely debilitated by the condition (Axelsson & Ringdahl, 1989). Yet, despite the prevalence and the fact that an estimated 20% of those affected will seek some form of treatment, currently this treatment focuses on coping with or covering it up with other sounds. The current clinical practice guidelines recommend education, hearing aid evaluations, cognitive behavioral therapy, and sound therapy (Tunkel et al., 2014). However, many investigators are looking to gain a greater understanding of the mechanisms underlying tinnitus in an effort to develop more effective and targeted interventions. One area that has received significant attention is the role that the central nervous system may play in this disorder. Tinnitus is no longer thought of as just a disorder of the ear or a phantom symptom, but rather as an abnormal auditory perception reflecting dysregulation in development or neural firing within the central nervous system (Baguley, 2002; Shulman, Avitable, & Goldstein, 2006). EEG studies, while not yet conclusive, are beginning to reveal some patterns. Abnormal EEG power has been reported in 67.2% of tinnitus subjects, with the greatest deviations in the delta and beta frequencies. Specifically, increased beta power in the frontal and temporal regions (anterior cingulate, amygdala, insula, and parahippocampus) has been correlated with reported severity (Shulman et al., 2006) and emotional distress (Meyer, Luethi, Neff, Langer, & Buchi, 2014; Vanneste & De Ridder, 2013). Maladaptive coping has been correlated with increased alpha power in the left frontal regions (dorsolateral prefrontal cortex and orbitofrontal), and anterior cingulate, and increased connectivity within the default mode network (Joos, Vanneste, & De Ridder, 2012; Vanneste, Joos, Langguth, Ting To, & De Ridder, 2014). Additionally, the perception of tinnitus has been associated with increased gamma range activity in the auditory cortex. Gamma activity is correlated with perceptual binding and conscious awareness; however, in the case of tinnitus, this activity may result in becoming aware of a stimulus that isn t really there (De Ridder & Van Der Loo, 2008; Meyer et al., 2014). Thus, tinnitus may result from dysregulation in several networks with abnormal gamma activity in the auditory cortex contributing the perception of tinnitus and dysregulation of the frontotemporal regions contributing to maladaptive coping, distress, and depression. In addition to dysregulation of the central nervous system, dysregulation of the autonomic nervous system (ANS) has

Meckley Kutyana also been proposed to play a role in tinnitus. Although the ANS has been largely ignored in research (De Ridder, Van Der Loo, & Coelho, 2006), clinically it is often observed that increased arousal and stress correlates with exacerbation of tinnitus. Autonomic regulation has been evaluated through the use of heart rate variability (HRV). Early research found a correlation between HRV and tinnitus, with HRV being decreased, particularly for individuals who reported significant stress (Datzov et al., 1999). Early research by Matsushima and colleagues in 1996, reported by Vanneste and De Ridder (2013), also found tinnitus suppression to be associated with increased parasympathetic tone. More recently, the neural correlates of HRV are beginning to be identified and while further research is needed to fully understand these relationships, it is interesting that brain regions that are thought to influence HRV (anterior cingulate, insula) have also been implicated in tinnitus distress (Van Der Loo, Congedo, Vanneste, Heyning, & De Ridder, 2011; Vanneste & De Ridder, 2013). This knowledge and expanded view of tinnitus has led to investigating other modalities that may target the dysregulated physiology of tinnitus more directly. The majority of these studies have utilized neurofeedback to regulate EEG activity, primarily in the alpha, beta, and/or delta frequencies. A positive impact on tinnitus perception and/or distress has been reported and changes in EEG activity have been observed in some, but not all, studies (Dohrmann, Weisz, Schlee, Hartmann, & Elbert, 2007; Gosepath, Nafe, Ziegler, & Mann, 2001; Schenk, Lamm, Gundel, & Ladwig, 2005). However, direct regulation of the autonomic nervous system has yet to be explored. Currently, only one study has been published investigating the use of vagal nerve stimulation (De Ridder, Vanneste, Engineer, & Kilgard, 2014) and there are no studies investigating the use of HRV biofeedback. Thus, the aim of this study is to explore the impact HRV biofeedback may have on tinnitus perception and distress through regulation of the autonomic nervous system. Methods Participants Three unpaid volunteers (2 female, 1 male) ages 50, 52, and 65 were recruited from Synchronicity Wellness, a concierge medical practice in Asheville, North Carolina. All participants denied a history of head trauma. Equipment Physiological measurements including heart rate (recorded by electrocardiogram and blood volume pulse sensors), respiration, finger temperature, and skin conductance were recorded using the NeXus-10 MKII (Mind Media BV, Herten, The Netherlands; http://www.mindmedia.info/). End tidal carbon dioxide (CO 2 ) was measured using a capnometer, the CapnoTrainer (Better Physiology, Ltd., Santa Fe, New Mexico; http://www.betterphysiology.com/). Interbeat intervals and measures of heart rate variability were calculated using BioTraceþ software from MindMedia BV. The interbeat interval (IBI) is the time in milliseconds between consecutive heart beats, in the case of the electrocardiogram, the distance between consecutive R waves. Artifacts were removed from the IBI record through manual and automatic procedures with the IBI being removed and interpolated if the difference between consecutive IBIs was greater than 25%. Procedure This preliminary investigation, modeled after the protocol by Leher et al. (2013), spanned a 6-week period. During the first week, the rationale was explained and informed consent was obtained. Participants then completed the Tinnitus Handicap Inventory (McCombe et al., 2001) and a standard audiology and tinnitus evaluation was performed. ECG, respiration, temperature, skin conductance, and capnography sensors were applied and participants sat quietly for 5 minutes with eyes open. Participants resonant frequency breathing rate was then determined using the Resonance Rate Determination protocol developed by Rosenthal and Dornheim (personal communication). In the second week, the participants returned for the first HRV biofeedback training session. Blood volume pulse, respiration, temperature, and skin conductance sensors were applied and participants viewed a screen (Figure 1) displaying their respiration as a ball that inflated as they inhaled and deflated as they exhaled. The screen also displayed a pacer set at their resonance frequency breath rate and they were instructed to breathe along with the pacer. Respiration rate and LF% power were also displayed. The session lasted for 20 minutes. They were instructed to practice at home using an online pacer. Participants then returned for three more weeks and the same procedure was followed. Following four training sessions, posttraining measures were taken in the same manner as described in week 1. Measures Heart rate variability. Heart rate variability was evaluated using both the frequency domain and time domain statistics. In the frequency domain, spectral analysis was used to assess the contribution of high frequency (0.15 0.5 Hz), low frequency (0.04 0.15 Hz), and very low frequency (0.01 Biofeedback Fall 2015 143

HRV Biofeedback for Tinnitus Figure 1. BioTraceþ software display, showing a paced breathing training screen. 0.04 Hz) components. In the time domain the SDNN, RMSSD, and pnn50 were used to evaluate variability. (Each of these indices is a statistical measure of heart rate variability. The SDNN is the standard deviation of the normalized interbeat interval. The RMSSD is the root mean square of successive differences in the interbeat interval. The pnn50 is the percentage of the interbeat intervals, which differ from adjacent intervals by 50 milliseconds or more). Tinnitus questionnaire. The distress produced by tinnitus was evaluated using the Tinnitus Handicap Inventory (THI). It consists of 25 questions under the three subscales of emotional, functional, and catastrophic response. The THI is scored as shown in Table 1. Individual Results Case 1 BS is a 50-year-old married female employed as a health care professional. She reported that her tinnitus has been present as long as she can remember but became pronounced 15 years ago following a surgery. It was attributed to medication and once she came off the medication the tinnitus improved but was still present. She rated her Table 1. Scores on the Tinnitus Handicap Inventory range from zero to 100, and are rated slight to catastrophic 144 G Score Description 1 0 16 Slight: Only heard in quiet environment, very easily masked. No interference with sleep or daily activities. 2 18 36 Mild: Easily masked by environmental sounds and easily forgotten with activities. May occasionally interfere with sleep but not daily activities. 3 38 56 Moderate: May be noticed, even in the presence of background or environmental noise, although daily activities may still be performed. 4 58 76 Severe: Almost always heard, rarely, if ever, masked. Leads to disturbed sleep pattern and can interfere with ability to carry out normal daily activities. Quiet activities affected adversely. 5 78 100 Catastrophic: Always heard, disturbed sleep patterns, difficulty with any activity.

Meckley Kutyana Table 2. HRV (SDNN, RMSSD, pnn50, LF% power), respiration rate, skin conductance (SC), finger temperature, and CO 2 data for each baseline and training session for Case 1 HRV Training Sessions Pretraining Baseline Week 1 Week 2 Week 3 Week 4 Posttraining Baseline SDNN 65.51 105.35 72.21 68.36 55.07 64.64 RMSSD 35.08 80.08 45.68 45.23 36.57 39.18 pnn50 12.61% 45.64% 24.89% 22.25% 13.23% 17.86% LF% power 91.81% 80.44% 88.51% 89.95% 91.28% 89.18% Resp rate 4.97 7.14 6.68 6.53 6.53 5.28 SC 1.46 2.51 2.07 1.38 2.74 1.01 Temp 75.45 77.68 80.1 85.35 89.05 85.36 CO 2 28.6 37.2 THI score 39 36 tinnitus in the moderate range at baseline and reported that it didn t typically interfere with her functioning, but was annoying. She was generally able to ignore it and only was aware of it when in a quiet room, during times of stress, and when going to sleep. Slow breathing was not a challenge for her; however, her CO 2 levels were found to be in the hypocapnic range at 28.6 Torr (1 Torr ¼ 1 mm of mercury) during the baseline recording (hypocapnia signifies that an individual exhibits a deficient level of carbon dioxide). Interestingly, BS also reported a history of Raynaud s, poor circulation, and her hand temperature was found to average 758F during the baseline recording, which may be related to, or exacerbated by, the low CO 2 levels. In the first HRV training session, BS was trained in resonance frequency paced breathing, and a capnometer was also used to monitor CO 2 levels. She was coached on decreasing the tidal volume of each breath and lengthening the exhale. This was found to normalize CO 2 levels. Over the course of the 4-week training, her hand temperature increased each week. When comparing pre versus post baseline recordings, her hand temperature after training was found to be ten degrees higher and her CO 2 levels posttraining were in the normal range. No significant changes were noted in HRV; however, HRV was found to be fairly healthy at baseline and she did not exhibit excessive levels of stress. BS rated her tinnitus distress at 39, in the moderate range, at baseline and slightly lower at 36, in the mild range, at posttraining. Table 2 and Figure 2 show the results for Case 1. Case 2 LB was a 52-year-old married female who reported severe tinnitus that was always present and drives me crazy. It began in 2002 after she worked in construction for a year and did not wear ear protection when using power tools. She reported that it was much worse when the environment was Figure 2. The displays show the spectral distribution of HRV for Case 1. (left) Pretraining and (right) posttraining. Biofeedback Fall 2015 145

HRV Biofeedback for Tinnitus Table 3. HRV (SDNN, RMSSD, pnn50, LF% power), respiration rate, skin conductance (SC), finger temperature, and CO 2 data for each baseline and training session for Case 2 HRV Training Sessions Pretraining Baseline Week 1 Week 2 Week 3 Week 4 Posttraining Baseline SDNN 50.74 53.89 71.21 53.66 92.1 74.01 RMSSD 34.65 31.94 39.46 28.17 49.14 47.12 pnn50 7.62% 9.08% 9.80% 5.42% 19.02% 20.85% LF% power 77.77% 89.31% 93.89% 92.23% 92.05% 84.27% Resp rate 4.09 5.95 5.52 5.72 5.62 7.06 SC 1.57 1.63 1.18 2.82 1.73 1.41 Temp 94.9 94.09 90.06 92.46 88.71 87.62 CO 2 39.3 THI score 67 65 quiet and she reported using background sound as a distraction. LB had no difficulty breathing with the pacer and was able to quickly breathe at her resonance frequency. She reported feeling more relaxed after HRV training, but did not practice the technique at home stating that she forgot or just couldn t find the time. Despite the lack of home practice, measures of HRV improved over the course of training. However, little change was noted in the perception of tinnitus or accompanying distress. Her self-report indicated only a 2-point decrease on the THI scale, leaving her distress in the severe range. Table 3 and Figure 3 show the results for Case 2. Case 3 JB was a 65-year-old male who reported that he had struggled with significant tinnitus for many years and rated his distress as catastrophic. He was retired and indicated that he experienced considerable stress on a daily basis. This appeared to be largely due to the recent death of his wife. He was also struggling with pain in his neck (he reported that his doctors thought this pain was due to genetic issues and not due to a specific injury). During the initial session, JB was completely unable to breathe along with the breathing pacer and as a result a valid resonance frequency could not be determined. Given his struggle with breathing, he was coached on diaphragmatic breathing and was asked to practice daily. During the first HRV training session, the breathing pacer was set at 10 bpm since his resonance frequency could not be determined and this was slightly slower than his baseline breathing. He struggled with breathing and was only able to maintain a rhythmic pattern for 2 minutes at a time. When his breathing became irregular, the session was paused and he was asked to observe the differences in his breathing. Over the next three training sessions, JB was able to decrease his rate of breathing and increase the length of time he could maintain a rhythmic 146 Figure 3. The displays show the spectral distribution of HRV for Case 2. (left) Pretraining and (right) posttraining.

Meckley Kutyana Table 4. HRV (SDNN, RMSSD, pnn50, LF% power), respiration rate, skin conductance (SC), finger temperature, and CO 2 data for each baseline and training session for Case 3 HRV Training Sessions Pretraining Baseline Week 1 Week 2 Week 3 Week 4 Posttraining Baseline SDNN 26.51 33.18 53.32 54.26 48.17 20.62 RMSSD 15.86 21.07 32.18 29.76 20.72 12.46 pnn50.99% 1.67% 3.75% 4.37% 2.82% 0% LF% power 59.29% 75.91% 81.79% 82.54% 81.81% 45.06% Resp rate 12.98 10.1 11.29 8.02 6.52 12.33 SC.56.52.63.65.42.41 Temp 88.64 95.18 94.13 91.32 93.5 92.27 CO 2 35.6 34.6 THI score 80 49 pattern. In the last week of training he was able to breathe consistently at 6 breaths per minute for 20 minutes. The most notable change during training was an improvement in the HRV power spectrum and an increase in the percentage of HRV power in the low frequency range (LF%). However, there were no changes when comparing the pre baseline to the post baseline in measures of HRV, as he was not able to maintain the slower breathing rate without the pacer. Most importantly, his self-report ratings of tinnitus did change significantly with his prebaseline rating at 80, in the catastrophic range, to a rating of 49, in the moderate range, at postbaseline. Table 4 and Figure 4 show the results for Case 3. Discussion In summary, interesting patterns were observed for each case. Case 1 exhibited very low finger temperature (758F) and below normal CO 2 (28.6 Torr) at baseline. Following the 4-week HRV training period, finger temperature averaged 858F and CO 2 was well within the normal range (37.2 Torr). HRV measures did not change, but were healthy at baseline, and THI scores decreased by 3 points, moving from the moderate to mild range. Case 2 showed considerable improvement in HRV measures; however, the THI scores decreased only by 2 points and they remained in the severe range. Case 3 was clearly the most stressed, had the poorest HRV at baseline, and reported the most distress. The THI scores were found to decrease by 31 points, moving from the catastrophic to moderate range. However, there was no improvement noted in HRV measures when comparing pre versus post baseline, but increased LF% power and improved organization of the power spectrum was noted during training sessions. Figure 4. The displays show the spectral distribution of HRV for Case 3. (left) Pretraining and (right) posttraining. Biofeedback Fall 2015 147

HRV Biofeedback for Tinnitus Thus, change in some manner was noted in all three cases; however, a larger study with a more homogenous tinnitus population is warranted to determine if HRV biofeedback can be an effective intervention or adjunctive intervention to tinnitus treatment. Such a study may benefit from including participants with the poorest HRV at baseline and more severe distress associated with tinnitus. References Axelsson, A., & Ringdahl, A. (1989). Tinnitus A study of its prevalence and characteristics. British Journal of Audiology, 23, 53 62. Baguley, D. M. (2002). Mechanisms of tinnitus. British Medical Bulletin, 63, 195 212. Datzov, E., Danev, S., Haralanov, H., Naidenova, V., Sachanska, T., & Savov, A. (1999). Tinnitus, heart rate variability, and some biochemical indicators. International Tinnitus Journal, 5(1), 20 23. De Ridder, D., & Van Der Loo, E. (2008, August). Finding a neural correlate for tinnitus. Tinnitus Today, August, 20 21. De Ridder, D., Van Der Loo, E., & Coelho, C. (2006). Tinnitus and the autonomic nervous system. Retrieved August 10, 2015, from http://www.tinnitusresearch.org/en/meetings/files2006/ De_Ridder-Tinnitus_and_the_autonomic_nervous_system.pdf De Ridder, D., Vanneste, S., Engineer, N., & Kilgard, M. (2013). Safety and efficacy of vagus nerve stimulation paired with tones for the treatment of tinnitus: A case series. Neuromodulation: Technology at the Neural Interface, 17, 170 179. doi:10.1111/ner.12127 Dohrmann, K., Weisz, N., Schlee, W., Hartmann, T., & Elbert, T. (2007). Neurofeedback for treating tinnitus. Progress in Brain Research, 166, 473 485. Gosepath, K., Nafe, B., Ziegler, E., & Mann, W. J. (2001). Neurofeedback in der Therapie des Tinnitus [Neurofeedback training as a therapy for tinnitus]. HNO, 49(1), 29 35. Joos, K., Vanneste S., & De Ridder, D. (2012). Disentangling depression and distress networks in the tinnitus brain. PLoS ONE, 7(7), e40544. Kochkin, S., Tyler, R., & Born, J. (2011). MarkeTrak VIII: Prevalence of tinnitus and efficacy of treatments, The Hearing Review, 18(12), 10 26. Leher, P., Vaschillo, B., Zucher, T., Graves, J., Katsamanis, M., Aviles, M., & Wamboldt, F. (2013). Protocol for heart rate variability biofeedback training. Biofeedback, 41(3), 98 109. McCombe, A., Baguely, D., Coles, R., McKenna, L., McKinney, C., & Windle-Taylor, P. (2001). Guidelines for the grading of tinnitus severity: The results a working group commissioned by the British Association of Otolaryngologists, Head and Neck Surgeons, 1999. Clinical Otolarynogology and Allied Sciences, 26, 388 393. Meyer, M., Luethi, M. S., Neff, P., Langer, N., & Buchi, S. (2014). Disentangling tinnitus distress and tinnitus presence by means of EEG power analysis. Neural Plasticity, 2014, 1 14. Schenk, S., Lamm, K., Gundel, H., & Ladwig, K. H. (2005). Effects of neurofeedback-based EEG alpha and EEG beta training in patients with chronically decompensated tinnitus. HNO, 53(1), 29 38. Shulman, A., Avitable, M. J., & Goldstein, B. (2006). Quantitative electroencephalography power analysis in subjective idiopathic tinnitus patients: A clinical paradigm shift in the understanding of tinnitus, an electrophysiological correlate. International Tinnitus Journal, 12(2), 121 132. Tunkel, D. E., Bauer, C. A., Sun, G. H., Rosenfeld, R. M., Chandrasekhar, S. S., Cunningham, E. R., Archer, S. M., et al. (2014). Clinical practice guideline: Tinnitus. Otolaryngology Head and Neck Surgery, 151(S1), 1 40. Van Der Loo, E., Congedo, M., Vanneste, S., Van De Heyning, P., & De Ridder, D. (2011). Insular lateralization in tinnitus distress. Autonomic Neuroscience, 165(2), 191 194. Vanneste, S., & De Ridder, D. (2013). Brain areas controlling heart rate variability in tinnitus and tinnitus-related distress. PLoS ONE, 8(3), e59728. Vanneste, S., Joos, K., Langguth, B., Ting To, W., & De Ridder, D. (2014). Neural correlates of maladaptive coping: An EEG-study in tinnitus patients. PLoS ONE, 9(2), e88253. Andrea Meckley Kutyana Correspondence: Andrea Meckley Kutyana, MA, BCN, QEEGT, email: biooptimized@gmail.com. 148