An exploratory clinical study to determine the utility of heart rate variability analysis in the assessment of dosha imbalance.

The present study is a comparison of the data of spectral analysis of heart rate variability with clinical evaluation of pathological state of doshas. The calculated cardiointervalography values are combined into three integral indexes, which according to the authors' opinion reflect the influence on heart rhythm of vata, pitta and kapha, the regulation systems of the body known as doshas in Ayurveda. Seven gross dosha imbalances were assessed to test the agreement between the two methods in this study. Heart Rate Variability (HRV) spectral data was collected from 42 participants to make the comparison with the clinical assessment of dosha imbalance. Clinical method of dosha assessment and method of calculating integral indexes by cardiointervalography data showed substantial agreement by Kappa coefficient statistic (k = 0.78) in assessment of gross dosha imbalance. The results of the data generated from this pilot study warrant further studies to rigorously validate the algorithms of HRV analysis in understanding dosha imbalance in Ayurvedic clinical practice and research settings.

Introduction

Ayurveda explains homeodynamics and health in terms of balance of the three doshas (regulatory systems), whereas disease is the outcome of the failure to maintain homeodynamics and is understood in terms of the imbalance of doshas. In order to formulate a specific treatment protocol, an Ayurvedic physician has to determine the exact nature of imbalance of the doshas and the specific substratum in which the imbalance manifests. To a great extent, imbalance of the doshas can be understood by careful analysis of clinical symptoms and signs. However, in many instances, it can be challenging to compute the imbalance in terms of the dominance and relative increase or decrease of the doshas. In ancient times, Ayurvedic physicians developed protocols for detailed examination of tongue, eyes, urine, faeces, skin and so on to measure imbalance of doshas. Subtle techniques like pulse diagnosis were introduced later on and eventually became the gold standard for diagnosing dosha imbalance.

HRV and imbalance of the doshas?

HRV is the result of the impact of the autonomous regulation on the heartbeat. Doshas are also autonomous regulators of body physiology. It is well known that 90% of our responses to the challenges of the external environment are determined by the autonomous nervous system. Travis and Wallace have discussed about the possible correlations between autonomous nervous system and the doshas [1]. Tyagi and Cohen have reviewed evidence indicating that physiological and psychological stress disrupts autonomic balance, which has long-term implications in a wide range of mental and physical illnesses [2]. Several studies have demonstrated that HRV is a useful tool in assessing pathology as well as treatment outcomes in many diseases. Masel et al. showed that HRV could be a surrogate marker for alleviation of cancer through pain and could also detect pain without active participation of the patients [3]. Kim et al. have demonstrated that perioperative HRV correlated with pre-operative depressed mood in patients with hepatic cancer [4]. Koszewicz et al. profiled autonomic dysfunctions in patients with primary brain tumor revealing sympathetic hyperactivity through HRV analysis [5]. Zhou et al. have reviewed the role of HRV in the prediction of survival in patients with cancer [6]. Taffe et al. have assessed post prandial HRV spectral analysis to differentiate overweight from normal weight adults [7]. Gupta et al. concluded that a significantly raised central vagal outflow and a concomitant significantly low central sympathetic efferent could be appreciated in asymptomatic asthmatic patients as compared to that in the control group by HRV spectral data analysis [8]. These studies indicate that HRV can vary in different diseases and also in specific stages of different diseases, which is also the case with the Tridoshas. This study aims to generate pilot data that will help in understanding the correlation between dosha imbalance and HRV patterns in patients diagnosed with different clinical conditions.

Previous studies exploring the correlation between HRV and doshas

Recently, Harupjit Singh from the Electrical and Instrumentation Engineering Department Thapar University, Patiala explored the relationship between three Ayurvedic doshas and HRV frequency bands by conducting a pilot study for his masters thesis [11]. His study involved only twenty-five patients and it achieved classification accuracy of 53% in case of the vata dosha and 70% in case of the pitta dosha.

Materials and methods

Study description

For the purpose of this study, forty-two patients with different Ayurvedic constitutions and pathology were randomly selected. All patients were attending the Amrita School of Ayurveda Hospital and signed a voluntary informed agreement to participate in the study.

Criteria for exclusion from the study included: medications for cardiac ailments influencing chronotropy of sinus node; multiple heart rhythm failures; divergence in data in verification of pathological dosha dominance between checklist of dosha dominance symptoms and expert assessment; HRV recording with many artefacts.

A checklist of dosha dominance symptoms compiled from the classical texts was used to make an exhaustive listing of the clinical symptomatology of the patient. The patients were then seen by a team of clinical experts who assessed the dosha dominance by performing an independent clinical assessment of the patient. Eleven participants with discrepancy in dosha dominance based on checklist and clinical assessment were excluded from the study and forty-two participants with congruence in dosha dominance based on checklist and clinical assessment were included for HRV analysis.

Researchers who filled the dosha dominance checklist, the clinical experts and the experts who recorded HRV spectral data were mutually blinded from the results of the independent assessments.

A VedaPulse hardware and software kit was used (manufactured by Biokvant LLC, Russia) for objective quantitative assessment of pathological dosha dominance using algorithms of HRV analysis. These algorithms were developed with inputs from Ayurvedic professionals with expertise in pulse diagnosis as well as pilot studies in a number of people.

Analysis of pathological dosha dominance included sequential actions, including: registration of biopotentials of the work of the heart for 5 min by placing electrodes on the wrists, software filtration of the signal with further receiving of HRV periodogram and analysis of cardiointervalogram using methods of spectral data analysis [9].

Frequency analysis of cardiointervalogram was done using discrete Fourier transform with overlapping. Periodogram was divided into three equal parts and for each of them a Fourier spectrum was calculated, after which these spectrums were averaged (see Fig. 2). Using the Fourier analysis, we estimate three spectral intervals: VLF – interval of very low frequencies (blue color); LF – interval of low frequencies (red color); HF – interval of high frequencies (green color). Horizontal axis shows frequencies in Hz, vertical axis shows spectrum power in ms2.

Fig. 2

VLF, LF and HF spectrums.

Practically the quantitative calculation of the dosha balance was done the following way: we calculated absolute values of the spectrum and rated them using 10-grade system of measurement, in which case the values of 2–4 points were considered a relatively standard interval which roughly corresponds with the average population value of the norm by absolute values of spectrum (see Fig. 3). If the normalized values of doshas in points were higher than the indicated interval and spectrums in VLF and LF intervals were dominating, it corresponded with–vata-pitta. If VLF and HF spectrums dominated, it corresponded with vata-kapha. If LF and HF spectrums dominated, it corresponded withpitta-kapha. If all three spectrums had a rising tendency, it corresponded with VPK. In Fig. 3, Yellow bar (Factor “V”) – total value of vata dosha; red bar (Factor P) – total value of pitta dosha; blue bar (Factor K) – total value of kapha. Corridor of norm 2–4 is highlighted in green.

Fig. 3

Balance of the doshas.

To detect frequency intervals, we use the following frames:

For VLF – (Time = 25–300 s, frequency = 0.04–0.0033 Hz)

For LF – (Time = 6.6–25 s, frequency = 0.15–0.04 Hz)

For HF – (Time = 2.5–6.6 s, frequency = 0.4–0.15 Hz)

A group of scientists under the leadership of Dr. Sorokin began to do comparative analysis of the HRV data with the clinical analysis of the dosha balance starting from 2006, the results of which were published. Later, similar algorithms for calculating doshas by HRV were offered in the works of Kelkar et al. [10] and Harupjit Singh [11], where they showed the connection of these intervals with doshas on the sample of several hundred participants. The difference of our work is that the power of spectrum in intervals was rated using a 10-grade system, which simplified the comparison with the scoring system of dosha calculation by clinical method. Normalizing of the absolute spectral power to a 10-grade scale is necessary to create relative values of the secondary indices – factor V, factor P, factor K, which reflect the dosha balance. The use of such normalizing eases the perception of the balance of these factors between them, which is hard when comparing the absolute values of the specter. This normalizing does not change the mathematic approach to the calculation, it only facilitates better visualization of the data.

Scales to assess dosha dominance

There are simple to complex scales to assess dosha dominance and simplest scale categorizing dosha dominance in seven categories was chosen for comparing diagnostic agreement in this study. The most complex scale of dosha dominance has sixty-three categories, which take into account the permutations and combinations of the relative increase and decrease of the doshas.

These seven categories cover the single, dual (samsarga) and triple (sannipāta) combinations of doshas that are routinely assessed in clinical practice. At the next level of complexity, distinction is made between vata-pitta and pitta-vata, as well as between vata-kapha and kapha-vata and between kapha-pitta and pitta-kapha. At even higher levels of complexity, the relative increase and decrease of the doshas are estimated. For example, vata increased, pitta decreased, kapha normal and so on and so forth.

Dosha assessment was made by clinical examination of study participants by experts, which was then compared with HRV spectral data. Algorithms based on HRV data were used to assess dosha dominance from ECG signals. By a mathematical analysis of high frequency, low frequency and very low frequency signals from the ECG readings, the dominant influence of the doshas on the regulatory systems of the body were calculated. Fig. 1, Fig. 2, Fig. 3 illustrate the capture of ECG signals and interpretations in terms of the doshic dominance by the HRV analysis. In Fig. 1, the upper graph shows the calculation of cardiac cycle duration in milliseconds via measuring the RR intervals of ECG. The lower graph shows the dependence of cardiac cycle duration variations (vertical axe, in ms) from the numerical order of cardiointerval (horizontal axe).

Fig. 1

Cardiac cycle duration variations.

Results

Table 1 gives the genderwise distribution of the study participants and it can be seen that nearly two-third were female. Table 1 gives the age group wise distribution of the study participants revealing that two-third were above the age of 50 and the maximum number of participants were in the age group of 61–70. The numbers in the tables indicate the frequency of participants in a specific category.

Table 1

Age and Gender wise distribution of patients.

Age group	21–30	31–40	41–50	51–60	61–70	71–80	81–90	Total
Male	1	2	2	3	6	1	1	16
Female	2	4	3	5	9	3	0	26
Total	3	6	5	8	15	4	1	42

Interestingly, both HRV analysis and clinical assessment did not identify extreme pitta or kapha imbalances. There were very few participants for the extreme vata, kapha-pitta and Tridosha imbalances. Majority of the participants were identified as presenting either vata-pitta or vata-kapha imbalances (see Table 4).

Table 4

Diagnostic agreement between HRV analysis and clinical assessment of dosha.

Clinical method	HRV analysis
	V	P	K	VP	VK	KP	VPK
V	2	0	0	0	0	0	0
P	0	0	0	0	0	0	0
K	0	0	0	0	0	0	0
VP	1	0	0	11	2	2	1
VK	2	0	0	4	14	1	1
KP	0	0	0	0	1	0	0
VPK	0	0	0	0	0	0	0

Table 3 lists the clinical diagnosis of the study participants who participated in the study.

Table 3

The clinical diagnosis of the study participants who participated in the study.

Allopathic diagnosis (number of participants)	Ayurvedic diagnosis
Brachial neuralgia (1)	Apabahukam
Cervical spondylosis (1)	Manyasthambham
Diabetes melittus (1)	Prameha
Gouty arthritis (2)	Vataraktam
Ligament tear (1)	Marmabhighatam
Low back ache (2)	Katigraha
Lumbar spondylosis (3)	Katigraha
Obesity (1)	Sthoulya
Osteoarthritis (11)	Sandhigata vata
Partial paralysis (1)	Vatavyadhi
Peripheral neuropathy (1)	Padasupthi
Psoriasis (1)	Sidhma kushtam
Rheumatoid arthritis (4)	Amavata
Sciatica (11)	Gridhrasi
Sinusitis (1)	Peenasa

Out of the forty-two participants who completed the study, twenty-seven (64.28%) showed complete agreement and ten (23.80%) showed partial agreement in assessment made by HRV analysis and clinical assessment. There was complete disagreement in five participants (11.90%). In the case of partial agreement, there was a mismatch in the assessment of the second dosha (see Table 4).

In order to assess the diagnostic agreement between the HRV analysis and the clinical assessment, the Kappa co-efficient was calculated using seven categories. The results are tabulated in Table 2. The values in the diagonal of the table indicate the number of participants classified by HRV analysis and clinical assessment into the same category. Values in other cells indicate number of participants in which disagreement was observed.

Table 2

Dosha vikriti categories of study participants.

	HRV analysis	Clinical assessment
V	5	2
P	0	0
K	0	0
VP	15	17
VK	17	21
KP	3	1
VPK	2	1

Kappa co-efficient is a statistical measure to assess inter-rater agreement more accurately than a simple percentage analysis. It is helpful to assess diagnostic tests that involve some degree of subjective interpretation by the observers. Its reliability lies in its ability to take into account the agreement occurring by chance. A Kappa score of 1 indicates perfect agreement and a score of 0 indicates complete disagreement. Even a score of 0.41–0.60 indicates moderate agreement, while a score of 0.61–0.80 indicates substantial agreement. A score of 0.81–0.99 indicates perfect agreement [12].

Analysis of the data gave Kappa statistic score of 0.78, which means substantial agreement between the HRV analysis and the clinical method of assessing doshic imbalances in pathological conditions.

Discussion

In the context of Ayurveda, doshas are the functional parameters which auto-regulate the physiology and therefore can have an impact on the heart rhythm.

Therefore, if there is a continuity between the formation of the pulse wave and the work of cardiac pacemakers, then instrumental analysis of HRV brings an important input to the understanding of physiological regulation processes, which connects disease development with pathological dosha dominance.

Modern technologies of biosignal registration and mathematical analysis refracted through the prism of empirical knowledge of Ayurvedic medicine allowed to develop an algorithm for assessing psychophysiological constitution and pathological dosha dominance. Analyzing spectral and temporal characteristics of HRV, a quantitative assessment system was developed to assess pathological dosha dominance, which was manifested in the VedaPulse hardware and software kit. This allowed building a connection between capabilities of modern electrophysiology and Ayurvedic medicine. The spectral analysis was carried out on the basis of the assumption that VLF, LF and HF grossly reflect vata, pitta and kapha respectively.

Assessment of the dosha imbalance is of cardinal importance in Ayurvedic diagnosis and treatment of diseases. However, the assessment is highly subjective and there is no validated instrument available for this purpose. There is a high inter-rater variability amongst experts in the field of Ayurveda. For this reason, only the gross dosha imbalances were selected for comparison in this study.

Examination of pulse is expected to give deeper insights into the state of doshic imbalance than mere clinical assessment. Therefore, clinical assessment is not the preferred standard to test the accuracy of HRV based algorithms in assessment of dosha imbalance. The advantage is that the clinical method of dosha assessment based on symptom analysis is more objective than pulse analysis by experts. Series of studies comparing HRV spectral analysis with clinical assessment of dosha dynamics could potentially help in identifying and characterizing HRV signatures that can predict imbalance of the dosha.

The present study has several limitations. HRV based algorithms used in the study to assess dosha imbalance are still in the early stages of testing and validation. Clinical assessment of doshas is not considered to be as accurate as assessment by pulse examination. An objective questionnaire for clinical assessment of dosha imbalance is not yet developed and validated. There is inter-rater variability amongst experts in clinical assessment of doshas. Matched control and adoption of non-linear mathematics and time domain analysis will be undertaken in future studies. Taking these limitations into consideration, only gross imbalance of doshas with consensus of experts were included for comparison with HRV based algorithms.

A larger multicentric study extending over a longer period that accounts for diagnostic error based on treatment outcomes will be able to make more accurate assessment of the scope of HRV analysis to determine imbalance of doshas.

Conclusions

Preliminary study to assess the diagnostic agreement in assessment of gross dosha imbalance by clinical method and HRV analysis showed substantial agreement as revealed by the Kappa score of 0.78.

The present study demonstrates the scope and relevance of further studies to validate the utility of the HRV in assessment of complex dosha imbalances and other parameters like dhatu imbalance.

Use of validated questionnaires for dosha analysis is recommended for future studies to assess diagnostic accuracy of HRV analysis. HRV analysis is not expected to replace the assessment of the physician by palpation and clinical examination. On the other hand, it could serve as a useful tool for research purposes and as an aid in the clinical practice of Ayurveda.

Sources of funding

The authors acknowledge the financial support received from Biokvant LLC, Russia for the conduct of the study.

Conflict of interest

The second author of the paper, Dr Oleg Sorokin is also sponsor of the study and the developer of the HRV analysis hardware device and the software tool used in the study (VedaPulse).