ABSTRACT This global study analyzed data from the largest dataset ever studied in the Heart Rate Variability (HRV) biofeedback field, comprising 1.8 million user sessions collected from

users of a mobile app during 2019 and 2020. We focused on HRV Coherence, which is linked to improved emotional stability and cognitive function. Positive emotions reported by users were

associated with higher Coherence scores and more stable HRV frequencies. In contrast, negative emotions exhibited lower scores and more dispersed frequency distributions. The most common

frequency associated with Coherence was identified at 0.10 Hz. However, many users with the highest levels of Coherence fell within a lower range from 0.04 to 0.10 Hz. Most users exhibited

frequency. The findings provide a strong foundation for future research and the development of targeted interventions aimed at enhancing emotional and physiological well-being through HRV

biofeedback and coherence practices. SIMILAR CONTENT BEING VIEWED BY OTHERS A META-ANALYSIS ON HEART RATE VARIABILITY BIOFEEDBACK AND DEPRESSIVE SYMPTOMS Article Open access 23 March 2021

EMOWEAR: WEARABLE PHYSIOLOGICAL AND MOTION DATASET FOR EMOTION RECOGNITION AND CONTEXT AWARENESS Article Open access 19 June 2024 POSITIVE AFFECT AND HEART RATE VARIABILITY: A DYNAMIC

ANALYSIS Article Open access 25 March 2024 INTRODUCTION Heart rate variability (HRV) biofeedback has emerged as a powerful tool for improving physiological and psychological well-being. Two

prominent approaches in this field are the HeartMath self-regulation techniques combined with real-time HRV coherence assessment and feedback technologies and the resonant frequency

breathing approach. Both methods aim to enhance aspects of HRV, but they differ in their specific focus and implementation1,2,3,4,5,6,7. Both approaches have demonstrated numerous benefits,

including reduced stress, and enhanced physiological function. They share a number of common outcomes such as increased HRV, improved autonomic function, lower blood pressure and better

overall well-being2,3,4,6,8,9,10,11,12,13,14,15. The research on resonant frequency breathing often emphasizes its use in clinical settings, particularly for treating various disorders and

health conditions1,16. This approach focuses on identifying and practicing breathing at an individual’s unique resonant frequency which results in the largest peak-to-trough amplitude in the

HRV waveform. This method aims to maximize efferent vagal modulation of the heart rhythm by increasing respiratory sinus arrhythmia and breathing at a rhythm near the resonant frequency

associated with the baroreflex system which is typically near 0.1 Hz, leading to increased HRV as compared to their resting state HRV. The typical protocol for this is to have participants

breathe following a pacer for 2 min at each of 5 frequencies: (6.5, 6, 5.5, 5, 4.5 breaths/min) for adults and 6.5 to 9.5 breaths/min for children. The range of the inter-beat-intervals

(IBI) is computed for each 2-minute record by looking at the peak-to-trough range (also called mean heart rate range). Once the breathing frequency resulting in the highest HRV amplitude is

determined, the individual is instructed to breath at that rhythm, typically with various types of breath pacers set to that frequency1. For an in-depth discussion of the HRV frequency

ranges and metrics see references17,18. In contrast, the HeartMath HRV coherence model seems to have gained more traction in educational, workplace, and consumer markets due to its

accessibility and broader focus on emotional self-regulation2,3,6,19. The HRV coherence model emphasizes the facilitation of self-regulation circuits and the neural interactions between the

heart and brain and the importance of the neural patterns and their stability in the afferent cardiovascular neural inputs to the brain on emotional stability and cognitive functions4,20.

Called the Heart-Rhythm Coherence hypothesis, it suggests that in addition to the amount of cardiovascular afferent vagal traffic to the brain, the patterns and stability of afferent inputs

modulate cognitive performance and self-regulatory capacity over both short (within one or two cardiac cycles) and longer time scales (minutes)4. Mather and Thayer propose a compelling

hypothesis suggesting that high-amplitude physiological oscillations have a causal impact on emotional well-being. Their theory posits that blood flow timing plays a crucial role in

determining brain network structure and function, with slower frequency, high-amplitude oscillations in heart rate potentially improving brain network dynamics. This improvement is

hypothesized to occur via fluctuations in blood flow, CO2 levels, and afferent inputs from breathing and the heart. These inputs are particularly likely to modulate activity in brain regions

associated with emotion regulation, especially prefrontal regulatory areas that are highly sensitive to physiological oscillations21. Supporting this perspective, several studies have

demonstrated that heartbeat-evoked potential amplitudes are significantly increased in numerous brain regions, including frontal areas, when participants engage in resonant frequency

breathing or coherence techniques. This hypothesis provides a potential mechanistic explanation for the observed relationship HRV and emotional well-being, suggesting that interventions

targeting physiological oscillations may have profound effects on brain function and emotional regulation22,23. The coherent state is characterized by a state specific higher amplitude,

smooth, sine-wave-like pattern in the heart rhythm that emerges when synchronization and frequency entrainment occur between the heart rhythms, blood pressure, and respiration rhythms. This

arises in the frequency range between 0.04 and 0.26 Hz of the HRV power spectrum, the coherence range3 in which there is increased synchronization between the brain rhythm and the cardiac

cycles4. The HRV coherence assessment algorithm is designed to detect the stability of the sine-wave like pattern in the heart rhythm independent of heart rate and the amplitude of the HRV

rhythm. This is done so it will work with individuals across the age span as the amount of HRV decreases with age24. It also needs to detect any frequency within the coherence range which

varies between individuals and can change within a session, tracking and feeding back a real-time coherence score. Therefore, the algorithm identifies the maximum peak in the 0.04–0.26 Hz

range of the HRV power spectrum, calculates the integral in a window 0.030 Hz wide, centered on the highest peak in that region, then calculates the total power of the entire spectrum over a

64-second window that is updated every 5-seconds. The Coherence ratio (CR) is formulated as: the product of two ratios: CR = (Peak Power/Total power below the peak frequency)*(Peak

Power/Total power above peak frequency) which is converted into a Coherence Score (CS) by taking the natural logarithm of CR + 1 which typically ranges between 0 and 82,4. The HRV coherence

feedback approach also includes the effect of emotional states on the heart rhythm. In earlier studies, we found that emotional states such as anger and appreciation are reflected

differently in the patterns of the heart rhythm and in the power spectra25. In past studies26,27 the focus was on the self-induction of various emotions and without any mention of altering

one’s breathing pattern. These studies found that participant’s breathing rhythms co-varied with the emotions they were experiencing suggesting that the brain structures above the

cardiorespiratory integration centers in the dorsal vagal complex involved in emotional processing and experience (amygdala, etc.) were unconsciously modifying the breathing rhythm, which is

reflected in changes in the heart rhythm4. Importantly, these studies found that positive emotions, such as appreciation and compassion were typically associated with a highly ordered heart

rhythm pattern, what is now called a coherent rhythm. Conversely, negative emotions, including anger, frustration, and anxiety, tend to produce irregular and erratic heart rhythm patterns.

These incoherent patterns manifest as jagged, disorganized waveforms, suggesting a lack of synchronized activity in the higher brain structures and between the two branches of the autonomic

nervous system28. Such states of incoherence have been linked to decreased physiological efficiency, increased energy expenditure, and potential long-term wear on bodily systems4,29. It has

also been demonstrated that analysis of the HRV pattern alone can be used to quantify discrete emotional states with around 75% accuracy30. Therefore, the heart rhythm coherence feedback

instruction and practice places greater emphasis on improving self-regulation and the importance of positive emotions as an important factor in sustaining heart rhythm coherence, which is

also referred to as heart-brain coherence, heart coherence, cardiac coherence, vascular system resonance and entrainment5. The instructions in the Inner Balance app and on-line instructional

videos for increasing heart rhythm coherence in stressful daily life contexts are given for two techniques. The Quick Coherence technique, which is designed to be used in-the-moment one may

be experiencing stressful feelings has two steps. The first step is called Heart Focused Breathing: Focus your attention in the area of the heart. Imagine your breath is flowing in and out

of your heart or chest area, breathing a little slower and deeper than usual. Find an easy rhythm that’s comfortable. The second step is: As you continue heart-focused breathing, make a

sincere attempt to experience a regenerative feeling such as appreciation or care for someone or something in your life. Try to re-experience the feeling you have for someone you love, a

pet, a special place, an accomplishment, etc., or focus on a feeling of calm or ease. The second technique, the Heart Lock-In is a heart-focused meditation technique that is used in longer

practice sessions. It shares the same first two steps but adds a third step which is to radiate that renewing feeling to yourself and others. Previous experience has shown that with this

approach individuals tend to naturally find their ideal breathing rhythm where the breathing rhythm, heart rhythm and beat-to-beat blood pressure rhythms frequency entrain at the resonant

frequency of the heart-brain-lung system25. We have consistently found that with this approach and a little practice, most people are able to use the slower and deeper breathing rhythm to

“jump-start” the shift into a more coherent rhythm and by then self-inducing a positive emotion, they tend to settle into their natural resonant frequency and sustain the coherent state for

longer periods compared to only using a paced breathing approach3. This approach may be more accessible for everyday use as it does not require determining a specific breathing frequency. It

also focuses on mental and emotional health and empowering individuals to self-regulate their thoughts, emotions and behaviors, which can have broader applications in daily life. Many of

the clinical, workplace and educational studies using HRV coherence feedback have included training in one or more self-regulation techniques designed to shift ones’ physiology into a more

coherent state when they are experiencing stress or wanting to enhance performance. Numerous studies that have used HRV coherence feedback technology to facilitate skill acquisition of

self-regulation techniques have found significantly improved key markers of health, wellness and performance in many health-care, law enforcement, corporate, military and educational

settings4,11,12,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50. In a series of novel HRV biofeedback studies that combined the two HRV biofeedback approaches, researchers first

determined participants’ resonant frequency and then used the coherence score, which better reflects heart rhythm stability, as the feedback signal for home practice sessions. A study with

younger and older adults found that after 5 weeks of daily HRV biofeedback practice that the intervention group had significantly increased cortical volume in left orbitofrontal cortex and

increased low frequency HRV power compared to the control group. These changes were significantly correlated with reductions in mood disturbance suggesting that daily biofeedback sessions

can increase resting state HRV and shape the brain circuits that help control HRV and regulate emotion51. Another study using the same intervention showed increased functional connectivity

between the medial prefrontal cortex and left amygdala. This was associated with increased down-regulation of activity in somatosensory brain regions during an emotion regulation task,

indicating that modulating heart rate oscillatory activity changes emotion network coordination in the brain52. A third study demonstrated that four weeks of practicing this intervention

produced large effect size reductions in plasma amyloid-β levels in the intervention group, while levels increased in the control group53. The most recent study in this series examined the

relationship between negative emotions and resting HRV. Pre-intervention assessments revealed significantly higher levels of negative emotions among participants with lower resting state

vagally mediated HRV. Post-intervention analysis demonstrated that participants exhibiting greater increases in resting HRV also showed more substantial decreases in negative emotions. A

mediation model incorporating all participants revealed that changes in resting HRV significantly mediated the relationship between training performance (i.e., coherent heart rate

oscillation during practice sessions) and changes in negative emotion. Notably, this mediation effect was significantly moderated by condition and was only significant in the intervention

group. These findings suggest that resting HRV changes mediated the mechanism by which coherent amplitude of heart rate oscillations reduced negative emotions, providing insight into the

physiological pathways through which HRV coherence training may modulate emotional states54. This study analyzes the largest dataset ever studied in the HRV biofeedback field, comprising

data from 1.8 million biofeedback sessions. It included a report of emotional states from a global population of individuals using the Inner Balance app (HeartMath, Boulder Creek, CA), who

uploaded their session data to a server called HeartCloud during the years 2019 and 2020. Past studies of Heart Coherence have been smaller controlled experiments, but this study explores it

in real-world settings and demonstrates its relevance in a broad global population. This study employed rigorous methodological approaches to ensure reproducibility and meaningful

contribution to HRV-related interventions. These methods allowed us to uncover novel insights into HRV patterns associated with eight different user-reported emotional states. We identified

the most prevalent coherence frequencies within this extensive dataset, examining their stability across various coherence levels and their modulation by emotional states. Our primary

objective was to provide new perspectives on HRV coherence feedback practices and their potential implications for enhancing self-regulatory capacity, health, and well-being. The analysis

focused on several key areas: (1) The identification of common coherence frequencies, (2) Assessment of frequency stability at different coherence levels, (3) Examination of emotional state

influences on HRV patterns, and (4) Evaluation of the relationship between coherence levels and reported emotional states. By analyzing this large-scale dataset, we aim to contribute to the

growing body of knowledge on HRV biofeedback and its applications. Additionally, we explore the importance of contextual factors in interpreting various HRV measures, emphasizing the need

for a nuanced understanding of HRV data in both research and clinical settings. This study’s findings have potential implications for refining HRV biofeedback techniques, improving

intervention strategies, and enhancing our understanding of the physiological correlates of emotional states. The results may inform future research directions and contribute to the

development of more effective HRV-based interventions for promoting health and well-being. RESULTS This study conducted a comprehensive analysis of Heart Rate Variability (HRV) data to

explore the relationship between heart rhythm coherence frequencies, emotional states, and physiological regulation. The Inter-Beat Interval (IBI) time series and Power Spectral Density

(PSD) analyses were utilized to demonstrate the relationship between coherence levels during HRV coherence feedback practice and HRV patterns. We compared average session coherence scores by

emotion to highlight the impact of positive and negative emotional states on physiological coherence. Additionally, we analyzed various time and frequency domain HRV metrics across

different coherence levels, revealing significant changes in autonomic nervous system activity at differing levels of heart rhythm coherence. The distribution of the most common coherence

frequencies and their associated breathing cycle periods were examined to underscore the importance of specific frequencies in achieving higher levels of coherence. Finally, we assessed the

variability in coherence frequencies by coherence level and emotion, emphasizing the increased stability associated with higher coherence levels and positive emotional states. These analyses

collectively provide a more comprehensive understanding of the role of heart rhythm coherence feedback and emotional states on physiological rhythms, enhancing self-regulatory capacity and

health outcomes. Figure 1 illustrates the relationship between coherence levels and the patterns of the heart rhythm and their corresponding Power Spectral Density (PSD) for three different

coherence levels: 0.7, 2.1, and 7.1. The top panel, with a coherence score of 0.7, shows a typical, irregular IBI pattern and a broad, low-amplitude PSD, indicating low coherence. The middle

panel, with a coherence score of 2.1, displays a more regular IBI pattern and a more pronounced peak in the LF region of the PSD. The bottom panel, with a coherence score of 7.1, exhibits a

highly regular, sinusoidal IBI pattern and a sharp, high-amplitude peak in the PSD around 0.1 Hz, reflecting high heart rhythm coherence. These results illustrate the relationship between

coherence levels and the regularity of the heart rhythm patterns and spectral characteristics of the HRV. To examine the relationship between emotional states and coherence levels during

heart rhythm coherence feedback practice sessions, we analyzed the average coherence scores by the reported emotion for sessions conducted globally in 2019 and 2020. Figure 2(a) illustrates

the average coherence scores by emotion for practice sessions uploaded to the server during this period, encompassing 1,884,216 user sessions of 3 to 15 min in duration. The sessions were

conducted globally, with the following distribution: North America (1,059,630 sessions), Europe (552,209 sessions), Asia (64,354 sessions), Oceania (61,923 sessions), South America (20,143

sessions), Africa (8,146 sessions) and other regions (117,811 sessions). Figure 2(a) reveals that positive emotions such as “Excited,” “Happy,” and “Peaceful” are associated with higher

average coherence scores (3.56, 3.49, and 3.40, respectively) compared to negative emotions like “Angry,” “Anxious,” “Bored,” and “Sad” (2.57, 2.83, 2.81, and 2.75, respectively). The t-test

results indicate that the differences in coherence scores between most emotions are statistically significant, with p-values of < 0.0001, except for “Content,” which shows no significant

difference (p-value: 0.28). Additionally, the overall comparison between positive and negative emotions shows a significant difference in average coherence scores, with positive emotions

yielding a higher average coherence score of 3.44 compared to 2.74 for negative emotions. The t-test conducted for the overall comparison between positive and negative emotions confirms this

significant difference, with a t-statistic of -207.84 and a p-value of < 0.0001 which strongly suggests that positive emotional states are associated with achieving higher coherence

levels during the feedback sessions. To provide a visual representation of the distribution of emotion counts among the user sessions, we analyzed the frequency of reported emotions. Figure 

2(b) provides a visual representation of the distribution of emotion counts among the user sessions. The pie chart shows that “Content” is the most frequently reported emotion after the

completion of a session, accounting for 30.1% of all sessions, followed by “Peaceful” (26.7%), “Happy” (21.9%) and “Excited” (6.1%). Negative emotions such as “Anxious” (6.2%), “Bored”

(3.6%), “Sad” (4.2%), and “Angry” (1.1%) make up a smaller portion of the overall distribution. This distribution indicates a predominance of positive emotional states among the users, which

aligns with the higher average coherence scores observed for these emotions in Fig. 2(a). The relatively high proportion of “Content,” “Peaceful,” and “Happy” suggest that users are more

likely to experience these positive states during the coherence sessions. To investigate how different coherence levels impact various HRV metrics, we conducted a detailed ANOVA analysis of

SDNN, Mean IBI, and natural logarithm of LF Power (LnLF), HF Power (LnHF), RMSSD (LnRMSSD), and LF/HF ratio (LnLF/HF), and the actual values are displayed in Fig. 3. Figure 3 shows the

results of the analysis of HRV metrics by coherence levels, revealing significant differences across these metrics. For SDNN, there is a clear decreasing trend from the 0–1 level (108.92 ms)

to the 3–4 level (88.63 ms), with a slight increase in the 5 + level (92.54 ms) which is due to the increased order in the distribution of the IBIs. Mean IBI, which is the inverse of heart

rate, shows a gradual increase from the 0–1 level (850.35 ms) to the 4–5 level (882.65 ms), with a slight decrease in the 5 + level (881.06 ms). As expected, LF Power increases significantly

from the 0–1 level (0.55 ms2) to the 5 + level (0.77 ms2), indicating an increase in the stability and amount of HRV. HF Power decreases from the 0–1 level (0.41 ms^2) to the 5 + level

(0.13 ms^2) due to the respiratory rhythms being frequency entrained with the heart rhythm in the LF region of the power spectrum. For RMSSD, there is a decreasing trend from the 0–1 level

(120 ms) to the 5 + level (80 ms), which would mistakenly be interpreted as reduced vagally-mediated HRV. However, this is not the case in this context as will be discussed in more detail in

the discussion section. The LF/HF Ratio shows a substantial increase from the 0–1 level (6.43) to the 5 + level (23.60), which in this context was expected as LF power was significantly

increased and is a signature of higher vagal activity and coherence levels. ANOVA results confirm significant differences in all metrics (_p_ < 0.0001), and post-hoc Tukey HSD (Honestly

Significant Difference), analysis used after ANOVA to find means that are significantly different from each other, revealed that these differences are significant across most pairwise

comparisons. For SDNN, the largest mean difference is observed between the 0–1 and 3–4 levels (-20.29 ms), while for Mean IBI, the largest difference is between the 0–1 and 3–4 levels (32.48

ms). LF Power shows the largest difference between the 0–1 and 5 + levels (0.22 ms2) and HF Power also shows the largest difference between the 0–1 and 5 + levels (-0.28 ms2). The LF/HF

Ratio exhibits the most substantial difference between the 0–1 and 5 + levels (17.17), highlighting the pronounced increase in LF power in the higher coherence levels. Taken together, these

findings suggest that higher coherence levels are associated with more stable and synchronized physiological states. To understand the distribution of the naturally occurring

coherence/resonance frequencies across the different levels of coherence, we analyzed the most common frequencies in the HRV spectrum among users’ sessions. Figure 4 provides a comprehensive

distribution of the most common coherence frequencies within a session among user sessions, involving a pool of 68,301 users. The data illustrates that the most common frequency observed is

0.1016 Hz, with 384,596 sessions having this frequency. Additionally, Fig. 4 also categorizes these frequencies by coherence levels, where it is evident that higher coherence scores (4–5

and 5+) have a greater association with the 0.1016 Hz frequency. Conversely, lower coherence scores (0–1 and 1–2) exhibit a more dispersed frequency distribution, although the 1–2 level has

the most sessions at 0.1016 Hz frequency. Surprisingly, a significant number of sessions had a coherence frequency below 0.1016 Hz, and the fraction of such sessions increased with coherence

scores. At the highest coherence scores (5+), the number of sessions at 0.1016 Hz dropped relative to lower scores while the number of lower-frequency sessions continued to rise. A

chi-squared test for independence further confirms that the association between coherence categories and frequency distributions is statistically significant, with a p-value of < 0.0001

indicating a strong dependence of frequency distributions on the coherence levels. These results imply that higher coherence levels are associated with a more stable and prevalent frequency,

around 0.1016 Hz for the majority of individuals but for some at even lower frequencies, especially at the highest coherence scores. To examine the relationship between HRV spectrum

frequencies and average coherence scores, we analyzed the average coherence scores across the range of coherence/resonance frequencies. Figure 5 shows the average coherence scores across the

range of frequencies in the HRV power spectrum. This analysis revealed that higher average coherence scores tend to be observed in the range of frequencies from 0.0469 to 0.1016 Hz. This is

particularly evident at low frequencies around 0.0547 Hz, 0.0625 Hz, and 0.0703 Hz, with coherence scores of 4.06, 4.05, and 4.05, respectively. These frequencies exhibit higher scores

compared to the 0.1016 Hz frequency, which has an average coherence score of 3.83. This suggests that while 0.1016 Hz is the most frequently observed coherence/resonant frequency, there is a

broader range of coherent frequencies that are being experienced, from 0.0469 Hz to 0.1016 Hz. It is also consistent with the observation that a larger fraction of users experience

frequencies below 0.1016 Hz at the highest coherence scores (5+). An ANOVA test further reinforces the finding that there are statistically significant differences in average coherence

scores across different frequency bins, with an F-statistic of 6893.01 and a p-value of < 0.0001. The relatively high average coherence scores at these lower frequencies may indicate that

these users are achieving a more stable and relaxed physiological state at these lower frequencies. This observation underscores the importance of exploring the full range of naturally

occurring heart rhythm frequencies to optimize feedback approaches to facilitate self-regulation and enhance mental, emotional and physiological well-being. To help visualize the inverse

relationship between breathing cycles in seconds and HRV spectrum frequencies, we include Fig. 6 which shows the conversion between the breathing cycle period and coherence frequencies. The

breathing cycle period for the most observed frequency, 0.1016 Hz, is approximately 9.84 s. Following this are the frequencies 0.0781 Hz and 0.0859 Hz, associated with breathing cycle

periods of approximately 12.81 s and 11.64 s, respectively. Notably, higher coherence levels tend to cluster around these specific lower coherence frequencies, suggesting that these are also

common naturally occurring resonant frequencies in many individuals. To explore how different emotional states may influence the heart rhythm, we analyzed the distribution of the HRV

frequencies by emotion. Figure 7 shows the distribution of the frequencies by emotion, revealing distinct patterns across different emotional states. Positive emotions such as “Peaceful,”

“Happy,” “Excited,” and “Content” show a pronounced peak around the 0.1016 Hz frequency, indicating a strong trend between positive emotions and this frequency. In contrast, negative

emotions like “Angry,” “Anxious,” “Bored,” and “Sad” exhibit a more dispersed distribution with less pronounced peaks, suggesting a less stable and consistent rhythmic frequency. The density

plots highlight that users experiencing positive emotions not only tend to have higher levels of coherence, they also have more stable frequencies, whereas those reporting negative emotions

display a broader range of frequencies. Overall, this distribution indicates that positive emotions are associated with more stable and harmonious physiological rhythms. To further

investigate the stability (variability) in HRV frequencies across different coherence levels and emotional states, we analyzed the standard deviation of coherence/resonance frequencies

within sessions and between the same users’ sessions. Figure 8(a) and (b) shows the frequency variability by coherence level and by emotion, respectively. Figure 8(a) shows the results of

the analysis of frequency variability revealing significant insights into the relationship between coherence levels and frequency stability. Sessions were segmented into 128-second intervals

with a 75% overlap, each with a coherence frequency, and summary statistics were calculated across all segments from all sessions. The mean and standard deviation in the frequency for each

coherence level shows a clear trend: as the coherence level increases, both the mean and standard deviation in the frequency decrease. Specifically, the mean standard deviation in coherence

frequency ranges from 0.0533 Hz in the 0–1 level to 0.0023 Hz in the 5 + level. Levene’s test55 for homogeneity of variances yields a test statistic of 110122.1668 with a p-value of < 

0.0001, indicating a significant difference in the standard deviations across coherence levels. This finding is further supported by the post-hoc Tukey HSD test, which shows significant

differences in mean standard deviations of frequencies between all pairs of coherence levels (p-adj < 0.0001) for all comparisons. The largest mean difference is observed between the 0–1

and 5 + levels (-0.051), while the smallest difference is between the 4–5 and 5 + levels (-0.0018). These results suggest that higher coherence levels are associated with more stable

frequencies in the HRV spectrum, as evidenced by lower means and standard deviations. The significant differences between all coherence levels highlight the robustness of this relationship,

indicating that as users achieve higher coherence, their heart rhythm pattern becomes more stable and consistent. This stability within sessions suggests that once participants settle into

their natural, individual coherence/resonant frequency, they are stable at this frequency. Figure 8(b) shows how the stability of the frequency varies significantly across different emotion

categories. Users who reported positive emotions such as “Peaceful,” “Happy,” “Excited,” and “Content” exhibited lower average standard deviations in their frequencies (0.0117 to 0.0130 Hz),

indicating higher stability. In contrast, users who reported negative emotions like “Angry,” “Anxious,” “Bored,” and “Sad” had higher average standard deviations (0.0164 to 0.0182 Hz),

reflecting less stability. Levene’s test and subsequent post-hoc Tukey HSD test confirm that these differences are statistically significant, with p-values of < 0.0001. This suggests that

individuals’ emotional states play a significant role in modulating the stability of the heart’s rhythmic pattern, with positive emotions being associated with more coherent and regulated

physiological states. The analysis of user data across sessions for the two years reveals that the majority of users exhibit high stability in their frequency from session to session. Taking

the standard deviation across common coherence frequencies from each session, most users have an average standard deviation below 0.012 Hz. This indicates that users generally maintain a

consistent heart rhythm frequency during their biofeedback sessions, suggesting that their unique natural coherence/resonant frequency remains the same across sessions. DISCUSSION The

results of our global study provide important insights into the relationship between the stability of the coherence/resonance frequencies and emotional states, as measured through Heart Rate

Variability data in Inner Balance app users in a real world setting. Our analysis, utilizing the largest dataset ever studied in the HRV biofeedback field, comprising data from over 1.8 

million sessions, uncovered several new key findings that enhance our understanding of heart rhythm coherence biofeedback and its implications for interventions to enhance emotional and

physiological well-being. First, the regularity in users naturally achieving a stable coherence frequency within and across sessions suggests that practice of the HeartMath emotional

regulation techniques combined with HRV coherence feedback are effective in inducing and sustaining a stable state of physiological coherence without the need to first determine one’s

resonant frequency. The spectral analysis of Inter-Beat Interval (IBI) data established that the most common frequency is around 0.10 Hz during practice sessions across all coherence scores

with the exception of the 0–1 level. This was not surprising as numerous studies have found this frequency is associated with the baroreflex system which regulates short-term blood changes56

and is at the center of the Coherence Range where frequency entrainment of the respiratory, beat-to-beat blood pressure and heart rhythms and increased heart-brain synchronization typically

occurs4,26. What was surprising is that for many of the more experienced users with the highest levels of heart rhythm coherence, their frequency was in the lower end of the Coherence Range

(see Fig. 4). In addition, as shown in Fig. 5, these lower frequencies have the highest overall coherence scores. Our analyses produced very strong p-values, which show that our findings

are reliable. The large dataset contributed to the strength of these results, reinforcing the validity of the statistical inferences drawn. The heart’s optimal function relies on a complex

interplay between the central nervous system and the intrinsic cardiac nervous system (ICNS) located within the heart itself. Neurocardiology research has extensively investigated the

anatomy and functions of the ICNS and its neural connections with the brain57. The intrinsic cardiac nervous system has both short-term and long-term memory functions, which can influence

HRV and afferent activity related to BP, rhythm, rate, and hormonal factors17. The ICNS exhibits both short-term and long-term memory capabilities, which can modulate heart rate variability

(HRV) and afferent activity associated with blood pressure, cardiac rhythm, heart rate, and hormonal factors. Comprising sensory, interconnecting, afferent, and motor neurons, the ICNS can

function autonomously from central neuronal control. The extensive network of these intrinsic cardiac neurons is so sophisticated that it has been characterized as a “little brain” within

the heart, highlighting its importance in cardiac regulation and adaptation58. The observation of higher coherence levels occurring at the lower frequencies may be related to the findings in

neurocardiology research that was undertaken after the observation in a auto-transplant study of dog hearts where the levels of HRV were much higher than expected suggesting that a

significant amount of HRV may be generated by the heart itself59. In long-term recordings from single afferent neurons in beating hearts and simultaneously from extrinsic cardiac neurons, it

was discovered that the lower frequency rhythms were generated from feedback loops between the sensory neurons in the heart and the intrinsic cardiac nervous system60. These rhythms

appeared to be associated with deeper states of relaxation and cardiovascular system efficiency as the stability and frequency of these low frequency rhythms were disrupted by stimulation of

efferent sympathetic inputs to the intrinsic cardiac nervous system17. As mentioned earlier, the coherence model emphasizes the importance of the neural patterns and their stability in the

afferent cardiovascular inputs to the brain on emotional stability and cognitive functions. This aspect of the model was informed by the seminal research and “baroreceptor hypothesis” first

introduced by psychophysiologists John and Beatrice Lacey who identified a causal relationship between the heart’s afferent inputs and cognitive performance61,62. This work was significantly

advanced by Wölk and Velden, who identified that it was not the amount or amplitude of the afferent inputs to the brain as the Lacey’s had thought, rather it is the rhythm pattern and its

stability that most important in influencing neurological functioning63,64. Studies in neurocardiology have since established that the interactions between the heart and brain are much more

complex than previously thought and that patterns of afferent activity occur over time scales ranging in milliseconds to minutes63,65. Heart-brain synchronization over shorter time periods

(seconds) was a key aspect of Wölk and Velden findings, while synchronization across much longer time periods (minutes) is an important aspect of The Heart Rhythm Coherence Hypothesis4. Our

study also revealed significant differences in coherence scores based on self-reported emotional states recorded at the end of each session. Positive emotions, such as “Excited,” “Happy,”

and “Peaceful,” were associated with higher coherence scores, reflecting more stable physiological states. Conversely, negative emotions like “Angry” and “Anxious” were linked to lower

coherence scores and more dispersed frequency distributions, indicative of less stable physiological states. These findings suggest that the self-activation of a positive emotion in step 2

of the techniques facilitates achieving and stabilizing a coherence state which can positively influence well-being, emotional stability and reduce mental and emotional stress. Analysis of

HRV metrics, such as SDNN, Mean IBI, LF Power, HF Power, RMSSD, and LF/HF Ratio, demonstrated significant changes across different coherence levels. The observed trends suggest a complex

interplay between respiration, beat-to-beat blood pressure regulation and activity in both the afferent and efferent activity in the autonomic nervous system. For instance, the expected

increase in the LF/HF ratio in higher coherence levels was due to large an increase in low-frequency power, which is often associated with enhanced baroreflex function and increased afferent

vagal traffic18,26,66. This increase in low-frequency power is typically due to an increase in the amplitude of the HRV rhythm and by a more regular sine-wave like rhythmic heart rate

pattern in the LF region of the spectrum4,26,67. However, the increased order in the heart rhythm results in a lower SDNN and RMSSD, despite either no change or an overall increase in the

amount (amplitude) of the HRV. This is because SDNN and RMSSD are more sensitive to capturing larger changes in the beat-to-beat interval variations, which are reduced in the more ordered

distribution of the inter-beat-intervals when the rhythm is more coherent68. Therefore, it is important to consider the context in which these measures are being interpreted and understand

that traditional interpretations of the SDNN and RMSSD do not capture the nuances of coherence or resonance states. In typical HRV analysis, SDNN and RMSSD values are often interpreted as

indicators of how much HRV occurred in a given analysis period with higher values being associated with greater amounts of overall HRV (SDNN) or increased parasympathetic activity (RMSSD)18.

However, during high coherence states, where the heart rate variability pattern becomes more regular and rhythmic, the increased regularity reduces the time between successive heartbeats,

leading to lower RMSSD values even when the overall HRV may actually be increased as indicated by an increased LF power. In a sine wave pattern, the differences between adjacent points are

smallest near the peaks and troughs, and largest at the midpoints between peaks and troughs. However, these differences change gradually and predictably. An example of this can be visually

seen in Fig. 1 where the IBIs are indicated by dots. In the highly coherent rhythm (1c), the time between consecutive pairs of heartbeats is clearly less than in the other less coherent

rhythms. As the RMSSD looks at the differences in the time intervals between consecutive heartbeats, the RMSSD is reduced in this context, although the amplitude of the rhythm is not.

Similarly, the overall variability measured by SDNN is reduced due to the more predictable and consistent heart rate pattern, with smaller beat-to-beat differences. In this context, the

lower RMSSD and SDNN does not imply reduced HRV or vagal activity but rather that different measures such as frequency domain measures are required to characterize these more ordered

coherent states. This is equally true in both resonant frequency breathing and HRV coherence contexts. In this context the coherence measure, LF power and the LF/HF ratio provide a more

appropriate measure as long as it is understood that the increased LF/HF ratio does not indicate increased sympathetic activity in this context. It is actually the opposite and indicates

increased vagal (parasympathetic) activity in this context18,26,66. Therefore, while SDNN and RMSSD are valuable metrics, their interpretation must be contextualized highlighting the need

for a more nuanced understanding in coherence research. The body of empirical evidence supporting the benefits of coherence-based interventions highlights the practical applications of HRV

coherence biofeedback to enhance self-regulatory capacity in both clinical settings and daily life across a broad range of applications such as law enforcement and military, classrooms,

corporate, and personal development contexts. By teaching techniques that increase coherence, combined with HRV feedback practice, these interventions have been shown to improve emotional

regulation, cognitive function, and overall well-being. The widespread availability of mobile applications like the Inner Balance Trainer and emWave Pro for clinical and educational

applications facilitates the integration of such practices into everyday routines, making coherence based self-regulation practices accessible to a global audience. However, there are

limitations to our study that need to be highlighted. This study is performed in real-world settings, which demonstrates its relevance in a broad global population, but it does not have the

same control as an experiment by design. Support is provided by reproducing results such as the 0.10 Hz coherence frequency and association with positive emotions, which is consistent with

past controlled studies4,25,26. It will be good to perform future controlled studies that test the new findings here. One significant limitation is the lack of pre-session emotion report

data, which means we do not have a reference baseline to compare to for any changes in emotional state or intensity of emotion due to the coherence practice of users. This limitation makes

it challenging to fully understand the impact of coherence practices on emotional states without knowing the initial emotional state of the participants. We also did not have any baseline

HRV data from the users before starting the practice session so we were not able to assess changes in users HRV, although this has been done in other studies4,25,32,37,47. Future research

could expand on these findings by exploring the long-term effects of coherence feedback practice and investigating the specific mechanisms through which coherence practices influence ANS

activity and cognitive functions. Additionally, studying the applications of coherence-based techniques in various clinical populations could provide deeper insights into their therapeutic

potential. Our study demonstrates that coherence-based interventions are not only beneficial for individual health but could also contribute to broader public health efforts aimed at stress

reduction and emotional well-being. With the increasing interest in HRV biofeedback and coherence practices, continued research in this field holds promise for advancing our understanding of

the intricate connections between physiology, emotion, and health. CONCLUSION This global study analyzed data from the largest dataset ever studied in the HRV biofeedback field, comprising

1.8 million user sessions collected from Inner Balance app users during 2019 and 2020. We focused on HRV Coherence, which is linked to improved emotional stability and cognitive function.

The most common frequency associated with Coherence was identified at 0.10 Hz. This is near the center of the Coherence Range where frequency entrainment of the respiratory, beat-to-beat

blood pressure and heart rhythms and increased heart-brain synchronization typically occurs. However, a surprising finding was that many users with the highest levels of Coherence fell

within a lower range from 0.04 to 0.10 Hz. Most users exhibited high stability (standard deviation < 0.012 Hz) in their coherence frequencies from session to session, and their stability

within a given session increases with increasing Coherence. The insights gained from this extensive dataset suggest that by instructing users to breathe deeper and slower and find a rhythm

that’s comfortable, they naturally find their unique resonant frequency. This supports the importance of techniques that instruct users to breathe deeper and slower than usual but find a

rhythm that is comfortable, allowing users to find their own individual HRV coherence or resonant frequency and sustain it. The observed trends in HRV metrics, such as SDNN, Mean IBI, LF

Power, HF Power, RMSSD, and LF/HF Ratio, suggest a complex interplay between coherent states and autonomic nervous system activity. Higher coherence levels are associated with more stable

and regulated physiological states, as evidenced by significant changes in these HRV metrics. It was also shown how and why it is important to consider the context in which these measures

are being interpreted and understood and that traditional interpretations of the SDNN and RMSSD do not capture the nuances of coherence or resonance states. This study also helps clarify the

relationship between common HRV coherence frequencies and emotional states by revealing that positive emotions are associated with higher coherence scores and more stable coherence

frequencies. In contrast, negative emotions exhibit more dispersed frequency distributions, indicating less stable physiological states. The insights gained from this extensive dataset

provide a strong foundation for future research and the development of targeted interventions aimed at enhancing emotional and physiological well-being through HRV biofeedback and coherence

practices. MATERIALS AND METHODS Anonymized HRV data from a global population of emWave Pro software and Inner Balance app users who elected to upload their session data to the HeartCloud

server was used in this study. The dataset for the years 2019 and 2020, comprised 70,041 unique users with a total of 4,637,295 HRV coherence practice sessions. Of these users, 43% were

female, 55% were male, and 2% did not specify their gender. The average age of participants was 49 years, with a standard deviation of 10.7 years. Out of the total sessions, 1,884,216 were

analyzed, focusing on those who also reported emotions and had durations of 3 to 15 min. The app interfaced with either a wired or Bluetooth-connected pulse photoplethysmogram (PPG) sensor,

which sampled at 125 Hz. We utilized established techniques for HRV analysis, including removal of artifacts, Welch’s method for Power Spectral Density (PSD) calculation, linear

interpolation for data resampling, and detrending procedures. The dataset included: UserUuid (a blinded unique identifier for each user), IBIStartTime (the start time of the Inter-Beat

Interval measurement in Unix timestamp), IBIEndTime (the end time of the Inter-Beat Interval measurement in Unix timestamp), LiveIBI (a string of comma-separated values representing time

intervals between heartbeats), AvgCoherence (the average coherence across the session), Country/Region (the user’s geographic location), and Emotion (self-reported user emotion following the

session). To begin, standardization of the Emotion column was carried out to consolidate non-English terms and merge similar categories. For example, all instances of ‘keine Stimmung’ and

‘Keine Stimmung’ were replaced with ‘No Emotion’, while terms such as ‘ängstlich’, ‘Ängstlich’, and ‘Gelangweilt’ were standardized to ‘Anxious’. Any records containing ‘No Emotion’ or with

AvgCoherence ≤ 0 were removed to maintain the data quality. Next, the number of sessions per unique user was calculated and a new column reflecting these counts was created. To ensure

session quality, only sessions lasting between 3 to 15 minutes were included; this range was chosen to better reflect users focused on a biofeedback session to ensure data integrity. The

UNIX timestamps (IBIStartTime and IBIEndTime) were converted to readable date-time formats. Subsequently, IBI data points that did not fall within the range of 400 to 2000 ms were excluded.

The LiveIBI data underwent validation to retain only non-empty entries. To facilitate analysis, the dataset records were further classified into predefined session AvgCoherence categories:

‘0–1’ (AvgCoherence greater than 0 and between 1 exclusive), ‘1–2’ (AvgCoherence between 1 inclusive and 2 exclusive), ‘2–3’ (AvgCoherence between 2 inclusive and 3 exclusive), ‘3–4’

(AvgCoherence between 3 inclusive and 4 exclusive), ‘4–5’ (AvgCoherence between 4 inclusive and 5 exclusive), and ‘5+’ (AvgCoherence ≥ 5). For coherence/resonant frequency analysis, the

initial 30 seconds of IBI data from each session were skipped to allow for the time it takes users to settle into a session to improve data consistency. The IBI data, initially recorded in

milliseconds, was converted to seconds for improved numerical stability. This data was then linearly interpolated to create a regularly spaced time series at 2 Hz, facilitating accurate

spectral analysis. The time series was segmented into windows of 128 s with a 75% overlap to maximize data utilization and reduce spectral leakage. To remove linear trends that could

introduce artificial low-frequency components, a detrending procedure was applied to the resampled IBI time series. Each segment was tapered using a Hanning window function to further reduce

spectral leakage at segment boundaries. Power Spectral Density (PSD) was calculated using Welch’s method. Within each 128-second window, the frequency corresponding to the peak PSD value

was identified as the coherence frequency confined within the physiological range of 0.03 to 0.4 Hz. This method provided robust estimates of the power distribution across different

frequency components of the time series. Various statistical measures for coherence frequencies—such as mean, standard deviation, mode, median, maximum, and minimum—were computed for each

session. Non-informative sessions, which lacked a mode or were bimodal, were removed from the analysis. Additionally, to derive further insights, the standard deviation of the most common

coherent frequency was calculated for each user across sessions and this was averaged across users, providing a comprehensive view of user coherence stability across sessions. Additionally,

the mean and maximum amplitudes of coherence frequency spectra were calculated. Finally, visualizations were generated to illustrate coherence frequencies by coherence level and various HRV

metrics, such as MeanIBI, SDNN, LF Power, HF Power, RMSSD, and LF/HF Ratio. Figures were also created to show common coherence frequencies by emotion, and the standard deviation in coherence

frequencies by coherence level and emotion. Given the large sample size in our study, the assumption of normality for the distribution of sample means is supported by the Central Limit

Theorem. This justifies the use of parametric tests such as t-tests and ANOVA. Additionally, all data were assumed to be independent and identically distributed, which is a prerequisite for

these statistical analyses. Our analyses were performed over very large datasets that produced strong p-values, which would hold true even if we reduced our degrees of freedom to accommodate

these assumptions. ETHICS STATEMENT This study was conducted as a secondary analysis of existing data, utilizing anonymized HRV biofeedback session data from a global population of users.

The analysis was performed in compliance with the privacy policy and data use agreement accepted by users of the HRV biofeedback application and was approved by the Kaunas University

Regional Ethics Committee for Biomedical Investigations (No.AMI_GML_21). Informed consent was obtained from all subjects and/or their legal guardian by their acceptance of the terms of

service and privacy policy of the HRV biofeedback application which grants use of anonymized data for research. This agreement clearly informs the user that anonymized data may be used for

research purposes and constitutes written informed consent. All personal identifiers were removed from the dataset prior to analysis, ensuring that individual users cannot be identified from

the research findings and strict confidentiality measures were maintained throughout the study to protect the privacy of the users whose data was analyzed. The aim of this research is to

contribute to the scientific understanding of HRV biofeedback and its potential benefits, with the ultimate goal of improving health and well-being for the broader population and the global

nature of the dataset ensures that the findings are representative of a diverse population, promoting equitable distribution of the potential benefits of this research. This research was

conducted in accordance with the Declaration of Helsinki and adheres to the ethical guidelines for human protection. DATA AVAILABILITY All datasets analyzed in the current study are

available from the corresponding author on reasonable request. REFERENCES * Shaffer, F. & Meehan, Z. M. A practical guide to resonance frequency assessment for heart rate variability

biofeedback. _Front. NeuroSci._ 14, 570400 (2020). Article  PubMed  PubMed Central  MATH  Google Scholar  * McCraty, R., Childre, D. & Coherence Bridging Personal, Social and Global

Health. _Altern. Ther. Health Med._ 16, 10–24 (2010). PubMed  Google Scholar  * McCraty, R. Following the rhythm of the heart: HeartMath Institute’s path to HRV Biofeedback. _Appl.

Psychophysiol. Biofeedback_, 1–12 (2022). * McCraty, R., Atkinson, M., Tomasino, D. & Bradley, R. The coherent heart: heart-brain interactions, psychophysiological coherence, and the

emergence of system-wide order. _Integr. Rev._ 5, 10–115 (2009). Google Scholar  * McCraty, R. & Childre, D. Coherence: bridging personal, social, and global health. _Altern. Ther.

Health Med._ 16, 10–24 (2010). PubMed  Google Scholar  * McCraty, R. & Zayas, M. Cardiac coherence, self-regulation, autonomic stability, and psychosocial well-being. _Front. Psychol._

5, 1–13. https://doi.org/10.3389/fpsyg.2014.01090 (2014). Article  Google Scholar  * Vaschillo, E. G., Vaschillo, B. & Lehrer, P. Characteristics of resonance in heart rate variability

stimulated by biofeedback. _Appl. Psychophysiol. Biofeedback_. 31, 129–142 (2006). Article  PubMed  Google Scholar  * Lehrer, P. M. & Gevirtz, R. Heart rate variability biofeedback: how

and why does it work? _Front. Psychol._ 5, 756 (2014). Article  PubMed  PubMed Central  MATH  Google Scholar  * Lehrer, P. M. et al. Biofeedback treatment for asthma. _Chest_ 126, 352–361.

https://doi.org/10.1378/chest.126.2.352 (2004). Article  PubMed  MATH  Google Scholar  * Goessl, V. C., Curtiss, J. E. & Hofmann, S. G. The effect of heart rate variability biofeedback

training on stress and anxiety: a meta-analysis. _Psychol. Med._ 47, 2578–2586 (2017). Article  CAS  PubMed  MATH  Google Scholar  * McCraty, R. & Atkinson, M. Resilience training

program reduces physiological and psychological stress in police officers. _Global Adv. Health Med._ 1, 44–66 (2012). Article  MATH  Google Scholar  * McCraty, R., Atkinson, M., Lipsenthal,

L. & Arguelles, L. New hope for correctional officers: an innovative program for reducing stress and health risks. _Appl. Psychophysiol. Biofeedback_. 34, 251–272.

https://doi.org/10.1007/s10484-009-9087-0 (2009). Article  PubMed  Google Scholar  * McCraty, R. & Tomasino, D. in _Clinical Psychology and Heart Disease Vol_. (eds Molinari, E.,

Compare, A. & Parati, G.). Springer, 488–509. (2006). * Bradley, R. T. et al. Reducing Test Anxiety and Improving Test Performance in America’s Schools: Results from the TestEdge

National Demonstration Study. (HeartMath Research Center, Institute of HeartMath, Publication No. 07-09-01, (2007). * Chaitanya, S., Datta, A., Bhandari, B. & Sharma, V. K. Effect of

resonance breathing on heart rate variability and cognitive functions in young adults: a randomised controlled study. _Cureus_ 14 (2022). * Bates, M. E., Lesnewich, L. M., Uhouse, S. G.,

Gohel, S. & Buckman, J. F. Resonance-paced breathing alters neural response to visual cues: proof-of-concept for a neuroscience-informed adjunct to addiction treatments. _Front.

Psychiatry_. 10, 624 (2019). Article  PubMed  PubMed Central  Google Scholar  * McCraty, R. & Shaffer, F. Heart Rate Variability: New perspectives on physiological mechanisms, Assessment

of Self-regulatory capacity, and Health Risk. _Glob. Adv. Health Med. Improv. Healthc. Outcomes Worldw._ 4, 46–61. https://doi.org/10.7453/gahmj.2014.073 (2015). Article  MATH  Google

Scholar  * Quigley, K. S. et al. Publication guidelines for human heart rate and heart rate variability studies in psychophysiology—part 1: physiological underpinnings and foundations of

measurement. _Psychophysiology_, e14604 (2024). * McCraty, R. New frontiers in Heart Rate Variability and Social Coherence Research: techniques, technologies, and Implications for Improving

Group Dynamics and outcomes. _Front. Public Health_. 5, 267 (2017). Article  PubMed  PubMed Central  Google Scholar  * Ayesha Ikhlaq, S. A. B. Muhammad Abul Hasan,. Phasic induction of

Bioelectromagnetic Heart-Brain Coupling through Emotional Stimuli. _J. Xi’an Shiyou Univ._ 19, 271–294 (2023). Google Scholar  * Mather, M. & Thayer, J. F. How heart rate variability

affects emotion regulation brain networks. _Curr. Opin. Behav. Sci._ 19, 98–104 (2018). Article  PubMed  PubMed Central  Google Scholar  * MacKinnon, S., Gevirtz, R., McCraty, R. &

Brown, M. Utilizing heartbeat evoked potentials to identify cardiac regulation of vagal afferents during emotion and resonant breathing. _Appl. Psychophysiol. Biofeedback_. 38, 241–255.

https://doi.org/10.1007/s10484-013-9226-5 (2013). Article  PubMed  Google Scholar  * Ikhlaq, A. et al. Phasic induction of Bioelectromagnetic Heart-Brain Coupling through Emotional Stimuli.

_J. Xi’an Shiyou Univ. Nat. Sci. Ed._ 19, 271–294 (2023). MATH  Google Scholar  * Umetani, K., Singer, D. H., McCraty, R. & Atkinson, M. Twenty-four hour time domain heart rate

variability and heart rate: relations to age and gender over nine decades. _J. Am. Coll. Cardiol._ 31, 593–601 (1998). Article  CAS  PubMed  Google Scholar  * McCraty, R., Atkinson, M.,

Tiller, W. A., Rein, G. & Watkins, A. D. The effects of emotions on short-term power spectrum analysis of heart rate variability. _Am. J. Cardiol._ 76, 1089–1093 (1995). Article  CAS 

PubMed  Google Scholar  * Tiller, W. A., McCraty, R. & Atkinson, M. Cardiac coherence: a new, noninvasive measure of autonomic nervous system order. _Altern. Ther. Health Med._ 2, 52–65

(1996). CAS  PubMed  Google Scholar  * McCraty, R., Atkinson, M. & Tiller, W. A. New electrophysiological correlates associated with intentional heart focus. _Subtle Energ_. 4, 251–268

(1995). Google Scholar  * McCraty, R. & Tomasino, D. in _In Stress in Health and Disease_. (eds Arnetz, B. B. & Ekman, R.). Wiley-VCH, 342–365. (2006). * Thayer, J. F., Åhs, F.,

Fredrikson, M., Sollers, I. I. I., Wager, T. & J. J. & D. A meta-analysis of heart rate variability and neuroimaging studies: implications for heart rate variability as a marker of

stress and health. _Neurosci. Biobehav. Rev_. 36, 747–756 (2012). Article  Google Scholar  * Leon, E., Clarke, G., Callaghan, V. & Doctor, F. Affect-aware behaviour modelling and control

inside an intelligent environment. _Pervasive Mob. Comput._ 6, 559–574 (2010). Article  Google Scholar  * McCraty, R., Atkinson, M. & Tomasino, D. Impact of a workplace stress reduction

program on blood pressure and emotional health in hypertensive employees. _J. Altern. Complement. Med._ 9, 355–369. https://doi.org/10.1089/107555303765551589 (2003). Article  PubMed 

Google Scholar  * Bradley, R. T. et al. Emotion self-regulation, psychophysiological coherence, and test anxiety: results from an experiment using electrophysiological measures. _Appl.

Psychophysiol. Biofeedback_. 35, 261–283. https://doi.org/10.1007/s10484-010-9134-x (2010). Article  PubMed  MATH  Google Scholar  * Li, W. C., Zhang, J., Kearney, P. & Braithwaite, G.

in _International Conference on Human-Computer Interaction._ Springer, 142–150. * Pyne, J. M. et al. Heart rate variability and cognitive bias feedback interventions to prevent

post-deployment PTSD: Results from a randomized controlled trial. _Mil. Med._ (2018). * Weltman, G., Lamon, J., Freedy, E. & Chartrand, D. Police Department Personnel stress resilience

training: an institutional case study. _Global Adv. Health Med_. 3, 72–79. https://doi.org/10.7453/gahmj.2014.015 (2014). Article  Google Scholar  * Lloyd, A., Brett, D. & Wesnes, K.

Coherence training improves cognitive functions and behavior in children with ADHD. _Altern. Ther. Health Med._ 16, 34–42 (2010). PubMed  Google Scholar  * Ginsberg, J. P., Berry, M. E.

& Powell, D. A. Cardiac coherence and PTSD in Combat Veterans. _Altern. Ther. Health Med._ 16, 52–60 (2010). PubMed  Google Scholar  * Field, L. et al. The tonic’s not always in a

bottle: a qualitative study investigating a heart rate variability biofeedback coherence intervention for individuals with gastrointestinal disorders. _J. Complex. Health Sci._ (2022). *

Field, L. et al. A systematic review of HeartMath© interventions to improve psychological outcomes in individuals with Psychiatric conditions. _Integr. Rev._ 17. (2021). * Trousselard, M. et

al. Cardiac coherence training to reduce anxiety in remitted Schizophrenia, a pilot study. _Appl. Psychophysiol. Biofeedback_, 1–9 (2015). * Jasubhai, S. Effect of quick coherence technique

on psychophysiological coherence, heart rate, stress, anxiety, Depression and feeling state in young adults in India. _J. Psychol. Neurosci._ 3, 1–9 (2021). Google Scholar  * Thurber, M.

R., Bodenhamer-Davis, E., Johnson, M., Chesky, K. & Chandler, C. K. Effects of Heart Rate Variability Coherence Biofeedback Training and Emotional Management techniques to decrease music

performance anxiety. _Biofeedback_ 38, 28–39 (2010). Article  Google Scholar  * Rijken, N. H. et al. Increasing performance of Professional Soccer players and Elite Track and Field athletes

with peak performance training and biofeedback: a pilot study. _Appl. Psychophysiol. Biofeedback_. 41, 421–430 (2016). Article  PubMed  PubMed Central  Google Scholar  * May, R. W.,

Seibert, G. S., Sanchez-Gonzalez, M. A. & Fincham, F. D. Self-regulatory biofeedback training: an intervention to reduce school burnout and improve cardiac functioning in college

students. _Stress_, 1–8 (2018). * Lemaire, J. B., Wallllace, J. E., Lewin, A. M., de Grood, J. & Schaefer, J. P. The effect of a biofeedback-based stress management tool on physician

stress: a randomized controlled clinical trial. _Open Med._ 5, 154–163 (2011). Google Scholar  * Alabdulgader, A. & Coherence A Novel Nonpharmacological modality for lowering blood

pressure in hypertensive patients. _Global Adv. Health Med_. 1, 54–62 (2012). Google Scholar  * Berry, M. E. et al. Non-pharmacological intervention for Chronic Pain in veterans: a pilot

study of Heart Rate Variability Biofeedback. _Global Adv. Health Med_. 3, 28–33. https://doi.org/10.7453/gahmj.2013.075 (2014). Article  MATH  Google Scholar  * Donghai, H., Wahab, M. N. A.

& Xiuling, Z. Effect of Heart Rate Variability Biofeedback Training on Score and Stress Level of Shooting Athletes. _Environ. Behav. Proc. J._ 9, 115–122 (2024). * McCraty, R. _Science

of the Heart Vol.2_. HeartMath Institute. ( 2016). * Ruiz, A. A. & Seco, M. M. Heart rate variability biofeedback intervention programme to improve attention in primary schools. (2024).

* Yoo, H. J. et al. Heart rate variability (HRV) changes and cortical volume changes in a randomized trial of five weeks of daily HRV biofeedback in younger and older adults. _Int. J.

Psychophysiol._ 181, 50–63 (2022). Article  PubMed  PubMed Central  MATH  Google Scholar  * Nashiro, K. et al. Increasing coordination and responsivity of emotion-related brain regions with

a heart rate variability biofeedback randomized trial. _Cogn. Affect. Behav. Neurosci._ 23, 66–83 (2023). Article  PubMed  MATH  Google Scholar  * Min, J. et al. Modulating heart rate

oscillation affects plasma amyloid beta and tau levels in younger and older adults. _Sci. Rep._ 13, 3967 (2023). Article  ADS  CAS  PubMed  PubMed Central  MATH  Google Scholar  * Jung, H.

et al. Changes in negative emotions across five weeks of HRV biofeedback intervention were mediated by changes in resting heart rate variability. _Appl. Psychophysiol. Biofeedback_, 1–24

(2024). * Gastwirth, J. L., Gel, Y. R. & Miao, W. The impact of Levene’s test of equality of variances on statistical theory and practice. _Stat. Sci._ 24, 343–360 (2009). Article 

MathSciNet  MATH  Google Scholar  * deBoer, R. W., Karemaker, J. M. & Strackee, J. Hemodynamic fluctuations and baroreflex sensitivity in humans: a beat-to-beat model. _Am. J. Physiol._

253, H680–689 (1987). CAS  PubMed  Google Scholar  * Armour, J. A. & Ardell, J. L. Oxford University Press, New York. (1994). * Armour, J. A. Potential clinical relevance of the ‘little

brain’ on the mammalian heart. _Exp. Physiol._ 93, 165–176. https://doi.org/10.1113/expphysiol.2007.041178 (2008). Article  ADS  CAS  PubMed  MATH  Google Scholar  * Murphy, D. A. et al. The

heart reinnervates after transplantation. _Ann. Thorac. Surg._ 69, 1769–1781 (2000). Article  CAS  PubMed  MATH  Google Scholar  * Kember, G. C., Fenton, G. A., Collier, K. & Armour, J.

A. Aperiodic stochastic resonance in a hysteretic population of cardiac neurons. _Phys. Rev. E_. 61, 1816–1824 (2000). Article  ADS  CAS  MATH  Google Scholar  * Lacey, B. C. & Lacey,

J. I. in _Cardiovascular Psychophysiology: Current Issues in Response Mechanisms, Biofeedback, and Methodology._ (eds P A Obrist, A H Black, J Brener, & L V DiCara). Aldine, 538–564.

(1974). * Lacey, J. I. & Lacey, B. C. Two-way communication between the heart and the brain: significance of time within the cardiac cycle. _Am. Psychol._, 99–113 (1978). * Wölk, C.

& Velden, M. In_ Psychobiology: Issues and Applications_. (ed Bond, N. W. & D. A. T. Siddle). Elsevier Science Publishers B.V., 371–379. (1989). * Wölk, C., Velden, M., Zimmerman, U.

& Krug, S. The interrelation between phasic blood pressure and heart rate changes in the context of the ‘baroreceptor hypothesis’. _J. Psychophysiol._ 3 (1989). * Wölk, C. & Velden,

M. Detection variability within the cardiac cycle: toward a revision of the ‘baroreceptor hypothesis’. _J. Psychophysiol._ 1, 61–65 (1987). Google Scholar  * Kromenacker, B. W., Sanova, A.

A., Marcus, F. I., Allen, J. J. & Lane, R. D. Vagal mediation of low-frequency heart rate variability during slow yogic breathing. _Psychosom. Med._ 80, 581–587 (2018). Article  PubMed 

Google Scholar  * Laborde, S. et al. Effects of voluntary slow breathing on heart rate and heart rate variability: a systematic review and a meta-analysis. _Neurosci. Biobehav. Rev._. 138,

104711 (2022). Article  CAS  MATH  Google Scholar  * You, M. et al. Influence of respiratory frequency of slow-paced breathing on vagally-mediated heart rate variability. _Appl.

Psychophysiol. Biofeedback_. 49, 133–143 (2024). Article  PubMed  MATH  Google Scholar  Download references AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * HeartMath Institute, Boulder Creek,

CA, 95006, USA Sai Balaji, Nachum Plonka, Mike Atkinson & Rollin McCraty * Capitol Technology University, Laurel, MD, 20708, USA Malathy Muthu * Faculty of Mathematics and Natural

Sciences, Kaunas University of Technology, LT-51368, Kaunas, Lithuania Minvydas Ragulskis * Institute of Cardiology, Lithuanian University of Health Sciences, LT-50162, Kaunas, Lithuania

Alfonsas Vainoras Authors * Sai Balaji View author publications You can also search for this author inPubMed Google Scholar * Nachum Plonka View author publications You can also search for

this author inPubMed Google Scholar * Mike Atkinson View author publications You can also search for this author inPubMed Google Scholar * Malathy Muthu View author publications You can also

search for this author inPubMed Google Scholar * Minvydas Ragulskis View author publications You can also search for this author inPubMed Google Scholar * Alfonsas Vainoras View author

publications You can also search for this author inPubMed Google Scholar * Rollin McCraty View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS

S.B., design of study, data and statistical analysis, writing; N.P., design of study and analyses, writing; M.A., organization of data collection, integration, and cleaning; M.M., MR, AV,

design of study and analyses, data visualization, writing; R.M., writing Introduction and discussion, coordinating data collection. CORRESPONDING AUTHOR Correspondence to Rollin McCraty.

ETHICS DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. ADDITIONAL INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional

claims in published maps and institutional affiliations. RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0

International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s)

and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material

derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a

credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted

use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/. Reprints and permissions

ABOUT THIS ARTICLE CITE THIS ARTICLE Balaji, S., Plonka, N., Atkinson, M. _et al._ Heart rate variability biofeedback in a global study of the most common coherence frequencies and the

impact of emotional states. _Sci Rep_ 15, 3241 (2025). https://doi.org/10.1038/s41598-025-87729-7 Download citation * Received: 30 October 2024 * Accepted: 21 January 2025 * Published: 25

January 2025 * DOI: https://doi.org/10.1038/s41598-025-87729-7 SHARE THIS ARTICLE Anyone you share the following link with will be able to read this content: Get shareable link Sorry, a

shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative KEYWORDS * Heart rate variability (HRV) *

Heart-rhythm coherence * Autonomic nervous system (ANS) * Emotional states * Biofeedback * Resonant frequency breathing