Neuro-cardiovascular correlates of controlled slow yogic breathing: A pilot study

Study design

This study is an interventional study of the before–after (pre– post) type, with a single arm and no comparator arm. The sole basis for concluding the study is the temporal relationship between the measurements done at 3 time points: pre (T₀), during (T₁), and post-ANB (T₂).

Subjects

Subjects willing to volunteer were recruited through the advertisement posted on the departmental notice board. Healthy male volunteers with no history of medical problems and no regular yoga practice were inducted into the study after obtaining written informed consent (n = 5; age 30 ± 3.8 years). The experimental protocol was explained, and ANB was taught. Ethical clearance was obtained from the Institute’s Ethics Committee before commencement (Ref. No. IESC/T156/30-03-2012).

Experimental protocol

Volunteers were asked to have a light meal at least 2 h before the recordings. Once reached, they were made to rest for 15 min in sitting. Then, electrocardiography (ECG) electrodes and 16 fNIRS optodes were placed, and the signals were checked. Other sensors, such as nasal canula for carbon dioxide and finometer for BP, were attached to the subject. The data for ECG, ETCO₂, beat-to-beat BP and brain oxygenation were recorded throughout the study period. The recording period spanned a total of 15 min (5 min each for pre-ANB [T₀], during ANB [T₁] and post-ANB [T₂]).

The manoeuvre of ANB

ANB is performed with one nostril closed at a time. The detailed procedure for doing ANB is described in the previous study.^[5]

Data collection and analysis

In the present study, Biopac MP150 (BIOPAC Systems, Inc., U.S.A.) with software ACQ version 4.2, Finometer (Finapres Medical Systems, The Netherlands), Capnograph (L & T, India) and fNIRS Model 1100-W (BIOPAC Systems, Inc., U.S.A) were used to acquire physiological signals. ECG100C and RSP100C amplifiers were used to simultaneously record lead II ECG and respiratory movements. Non-invasive beat-to-beat BP recording was done using the Finometer. All signals were acquired at a sampling frequency of 1KHz. Capnograph was used for ETCO₂ measurements, and the output signal was also received simultaneously on the ACQ platform. fNIRS was used to acquire 16 channels of continuous wave fNIRS signals. Details of fNIRS imaging are described elsewhere.^[6] Cognitive Optical Brain Imaging Studio software was used for fNIRS signal acquisition. Software ‘fNIRS Soft’ was used to evaluate prefrontal hemodynamic changes. The sampling rate was 2 Hz, and it gave 600 values of oxy-Hb, deoxy-Hb, HbT and oxygenation for the recording of T₀, T₁ and T₂ (300 s each).

The HRV, BPV and BRS were calculated using Nevrokard™ software version 3.2.0 (Nevrokard Kiauta d.o.o, Izola, Slovenia). Only time domain parameters of HRV (standard deviation of NN intervals [SDNN], NN50 count divided by the total number of all NN intervals [pNN50] ×100, the square root of the mean of the sum of the squares of differences between adjacent NN intervals [RMSSD]) and BPV (SDNN and RMSSD) were considered for analysis as frequency domain parameters were interfered with by constant respiratory frequency.

Statistical analysis

Data were screened for normality assumption using the Kolmogorov–Smirnov test and repeated measure analysis of variance was applied to compare means with post hoc analysis using GraphPad Prism 5.0 (GraphPad Software Inc., CA 92037 USA). P ≤ 0.05 was considered significant.

fNIRS analysis

We analysed the obtained data in the following ways to better understand the observations and decipher them based on temporal and spatial outcomes [Table 1].

Close

Figure 1:

(a) shows more increase in Oxy haemoglobin (Oxy Hb) at Left and Lower prefrontal cortex during and post alternate nostril breathing (ANB). (b) Shows significant and gradual increase in Oxy Hb with the increased duration of practice. (c) Frontal hemodynamic change at optode 8: Shows cerebral activation: Total haemoglobin, Oxy Hb increased and De-oxy-haemoglobin decreased with increased practice of ANB. ANB: Alternate nostril breathing, PFC: Prefrontal cortex, oxy-Hb: Oxy-haemoglobin, deoxy-Hb: Deoxy-haemoglobin, HbT: Total haemoglobin.

Export to PPT

1. The average values of oxy-Hb, deoxy-Hb and HbT were compared for T₀, T₁ and T₂ in the PFC region (1–16 optodes). Between T₀, T₁ and T₂, the values of oxy-Hb, deoxy-Hb and HbT were not significantly different. Further analysis was performed for the right, left, upper and lower regions of PFC. The left PFC is represented by optodes 1–8, while optodes 9–16 represent the right PFC. Upper PFC and lower PFC regions were represented by odd (optodes 1, 3, 5, 7, 9, 11, 13 and 15) and even (optodes 2, 4, 6, 8, 10, 12, 14 and 16) numbers. The percentage change was calculated for T₁ and T₂ with respect to T₀. Oxy-Hb and HbT levels increased significantly in T₁ and T₂ [Table 1]. The rise was greater on the left than on the right PFC and in the lower PFC than in the upper, as shown by the high percentage change values in [Figure 1a].

2. The paired t-test showed a significant increase in oxy-Hb during ANB at optodes 2, 4, 6 and 8 as compared to baseline. Further minute-to-minute changes in oxy-Hb concentrations were calculated and plotted for lower left PFC optodes 2, 4, 6 and 8. [Figure 1b] revealed that oxy-Hb increased gradually with the increased number of cycles of ANB.

3. Sequential changes in oxy-Hb, de-oxy Hb and HbT concentrations were plotted at optode 8 for T₀, T₁ and T₂ (averaged for 1 min epochs for 5 min duration each). [Figure 1c] further establishes a gradual increase in oxy-Hb and HbT, as well as a simultaneous decrease in deoxy-Hb at optode 8 at T₁ and T₂. Also, note that the T₂ values of oxy-Hb and HbT were higher as compared to T₀.

Cardiovascular oscillations

The HRV analysis indicated that during ANB (T₁) heart rate, SDNN, pNN50 and RMSSD were increased significantly, whereas SDNN increased and pNN50 and RMSSD were decreased significantly in T₂ when compared with T₀. A similar finding in BPV analysis was observed; the SDNN and RMSSD of SBP, DBP, and mean arterial blood pressure (MBP) were increased significantly during ANB. No significant change has been observed in recovery. We observed that the SBP decreased significantly in T₂ with respect to T₀, and MBP decreased significantly in T₂ than T₁ values, though the change in SBP and MBP was not significant during the T₁ [Table 2].

The BRS using the Sequence method showed an increasing pattern during ANB compared to the baseline. During T₁, the number of sequences for Up BRS decreased, whereas sequences for Down BRS converged to a narrow range [Figure 2].

Close

Figure 2:

(a-d) Showing an increasing pattern in Up-Baroreflex sensitivity (BRS) values and Up-BRS sequences decreased during-alternate nostril breathing (ANB). Down-BRS values did not change whereas down-BRS sequences were narrowed to an optimum value during-ANB. RRl: R peak to R peak duration from ECG recording in seconds, SBP: Systolic Blood Pressure

Export to PPT

ETCO₂

The overall findings of average ETCO₂ in T₀ (38.82 ± 2.61), T₁ (37.29 ± 7.04) and T₂ (37.2 ± 2.43) showed no significant change. However, when minute-to-minute values were compared, we found differential responses during ANB. The ETCO₂ of the three volunteers (S2, S4 and S5 depicted in green) was increased, whereas it was decreased in two (S1 and S3 depicted in red) in T₁, as shown in [Figure 3a].

Close

Figure 3:

(a-f) Showing end-tidal carbon dioxide, RR interval, mean blood pressure, oxy haemoglobin, deoxy haemoglobin and total haemoglobin pre-alternate nostril breathing (ANB), during-ANB and post-ANB. RR interval: R peak to R peak duration from ECG recording in seconds

Export to PPT

Correlation of the hemodynamic parameters at PFC and cardiovascular oscillations with respect to ETCO₂

The differential responses of ETCO₂ led us to analyse all the other parameters (in minute-to-minute values) for five volunteers individually and look for any pattern in the data. Although S1 and S3, who showed a decline in ETCO₂ on manoeuvre, had very different R-R intervals, their MBP showed few perturbations as compared to the other volunteers, as shown in [Figure 3b and c].

We also observed a reflection of this finding in the other recorded parameters. The average values of all 16 optodes were plotted for fNIRS data, we observed that the oxyHb remained lower than the mean in S1 and S3; also, the deoxyHb decreased during ANB. In S2, S4 and S5, the oxyHB increased during ANB and remained higher in recovery, as shown in [Figure 3d-f].

Effect of slow breathing on vmPFC

A recent study investigated the effects of a 1-month training program of Bhastrika pranayama (fast, forceful and deep breathing) on anxiety and emotion processing using functional magnetic resonance imaging. The researchers observed a notable decrease in anxiety and an increase in positive emotions. These changes were linked to the functioning and communication of a specific brain network that is responsible for processing emotions, including the amygdala, anterior insula, ACC, vmPFC, ventrolateral PFC and dorsolateral PFC.^[7] In the present study, the increase in oxy-Hb and HbT was relatively more in the lower left side of PFC that represents left vmPFC, indicating ANB may play a role in autonomic and affective regulation.

Mechanical effects of controlled slow yogic breathing

Multi-regulatory mechanisms are induced due to slow controlled yogic breathing through oscillations in cardiorespiratory and neurovascular systems.^[8] A transient increase in cerebral blood flow (CBF) during and after ANB could be attributed to increased slow cyclic oscillations in heart rate and BP. We observed the morphology of the signal and found large sinusoidal waveform pattern in ECG, BP and fNIRS signals during ANB. This suggests increased variability and seems to be of respiratory origin as they are paralleled with the fluctuations of respiratory movements. Possibly in the present study, at a gross level, it seems increased oscillations in heart rate, BP and CBF are due to the mechanical effect of slow breathing. It indicates that slow yogic breathing may have influence on central haemodynamics. Our previous publication has suggested that ANB triggers cardiovascular fluctuations by engaging multiple regulatory mechanisms. These mechanisms primarily involve mechanical coupling with breathing, specifically through central and nonbaroreflex mechanisms.^[5]

Effect on cardiovascular oscillations

The respiratory phase-driven fluctuations are known to modulate cardiovascular and cardiorespiratory coupling through synchronised hemodynamic fluctuations.^[9,10] In the present study, during ANB significant increase in SDNN, RMSSD indices of HRV and BPV indicate increased oscillations in R-R interval, systolic, diastolic and mean arterial BP while maintaining the baseline values [Table 2]. Increased BP oscillations activate arterial baroreceptors, which fire signals through afferent nerves to the cardiovascular centres in the medulla oblongata that regulate sympathovagal balance. In agreement with this fact in the present study, BRS and Up sequences showed an increased pattern simultaneously. Down BRS sequences narrowed to an optimum value during ANB [Figure 2]. It suggests that ANB enhanced the central reflex globally by utilising the baroreflex loop, improved oxygenation, heightened parasympathetic activity and BRS, in accordance with previously published studies.^[5,11,12] A significant increase in oxy-Hb at PFC may be attributed to the increased oscillations in MBP induced by ANB. This is further supported by a decrease in Up BRS sequences and an increase in down sequences, with a non-significant increase in BRS, indicating that the autoregulation has been maintained.

Effect of respiratory modulations in ETCO₂ (Chemoreflex sensitivity)

The differential autonomic modulation with enhanced BRS immediately after yogic breathing is reported in the literature.^[13] Similarly, the differential effect of ANB on ETCO₂ is observed in the present study, which is reflected in all the other haemodynamic and autonomic parameters [Figure 3]. The increased CBF may be due to repeated episodes of breath holding after inspiration as well as expiration as ANB was performed in the ratio of 4:2:4:2: Inhalation: Pause: Exhalation: Pause in seconds. This breathing pattern may pose a respiratory challenge, but vasodilatation of the cerebral arteries counterbalances a drop in perfusion pressure, maintaining adequate CBF in healthy individuals under normal conditions.^[14]

It is known that neuronal activation triggers an increase in CBF; hence, one of the factors for increased cerebral blood during ANB may be neural activation or vice versa. This can be warranted with simultaneous acquisition of EEG and fNIRS in future studies.