Cardiorespiratory synchronization in humans under paced respiration.

Noninvasive (not involving surgery) control of heart rhythm is very important for treatment of cardio-vascular diseases, e.g. various pathological arrhythmias. Most of the existing methods involve the use of medicines. However, fortunately, the heart rhythm can be changed not only by drugs. There are two aspects of heart rhythm that are important: the average heart rate and the regularity of heartbeats. It has been known for a long time that heartbeats are influenced by breathing: inhalation speeds the heart up an exhalation slows it down, which means modulation of inter-beat intervals. Modulation is closely related (but not equivalent to!) a phenomenon called synchronization (an introduction to the theory of synchronization phenomena could be found here). Recently it was shown that breathing can not only modulate, but also synchronize the heart rate: e.g. 1 breathing cycle can be made to correspond to exactly 3 heartbeats. Normally, breathing is irregular and synchronization very seldom occurs. One might suggest that if the breathing were regular and the heartbeats synchronized to it, they could become more ordered. It is well known that humans possess a remarkable ability to deliberately control their breathing: they can inhale and exhale at the required time moments. Thus, one can think of cardiorespiratory synchronization as of a promising tool for control of heart rhythm (another method is described here).

The easiest method to set the required breathing rate is via paced respiration, when a subject inhales when some external signal occurs (e.g. a sound pulse or a light blink). Paced respiration was studied in a number of works, however, there have been no systematic studies of heart rate synchronization by paced respiration, and a number of important questions remain open. For example, it is important to know whether cardiorespiratory synchronization is stable under variation of the breathing frequency. Does the phenomenon occur randomly, or are there ranges of respiration frequency when synchronization more probable, in spite of noise and nonstationarity? What types of n:m synchronization appear most frequently during paced respiration? The present work is an attempt to answer these questions.

Description of experiments
Measurements
Six volunteers participated in the measurements. From each of them an electrocardiogram (signal reflecting electrical activity of heart) and a respiratory effort (reflecting changes in thoracic or abdominal circumference that occur as the subject breathes) were recorded.

Scheme of measurements: endings of red lines on the human body show the positions of electrodes for ECG measurements; yellow line shows position of respiration transducer.		Photo of respiratory effort transducer by Biopac Systems Inc © used in experiments.

Typical electorcardiogram:		Typical signal of respiration:

Experiments:
Each subject was asked to lie on a bed in a comfortable position, with his head on a pillow, slightly tilted. An interval of about 20 min followed in order to achieve full relaxation. Values of heart rate were measured and compared several times during this period, to establish that relaxation had taken place. A signal with which to synchronize respiration was generated by the computer and presented to the subject as a series of light and sound pulses of duration 0.1 sec. The volunteer was able to see a red square on the computer screen and to hear the sound. The subject was asked to inhale when the sound and light signal appeared, and to exhale at his convenience before being prompted to inhale again by the next signal (see Figure on the left). We did not impose any restrictions on the amplitude of breathing, and suggested to subjects that they should breathe with whatever depth they felt most comfortable at the given frequency. After about 20 sec, the subject was breathing as prescribed by the generated signal and had become accustomed to the regime being imposed. Simultaneous measurements of the ECG and the respiration signals were then initiated, and continued for 180 sec. For each new measurement, the respiration frequency f_resp was increased by 1 breath per min ~i.e., by 1 min^-1). The minimal and maximal respiration frequencies were selected experimentally to be 3 and 30 min^-1, respectively, so that they should not be too difficult for the subject. Thus, 28 dataset pairs were measured for each subject. In between measurements the subject was allowed to relax for about 5 min until his heart rate, averaged over 1 min, had stopped changing. The average heart rate at rest was recorded to provide the current value of f₀before each new measurement. f₀will be used below to plot synchronization tongues.

Experiments:

Each subject was asked to lie on a bed in a comfortable position, with his head on a pillow, slightly tilted. An interval of about 20 min followed in order to achieve full relaxation. Values of heart rate were measured and compared several times during this period, to establish that relaxation had taken place. A signal with which to synchronize respiration was generated by the computer and presented to the subject as a series of light and sound pulses of duration 0.1 sec. The volunteer was able to see a red square on the computer screen and to hear the sound. The subject was asked to inhale when the sound and light signal appeared, and to exhale at his convenience before being prompted to inhale again by the next signal (see Figure on the left). We did not impose any restrictions on the amplitude of breathing, and suggested to subjects that they should breathe with whatever depth they felt most comfortable at the given frequency. After about 20 sec, the subject was breathing as prescribed by the generated signal and had become accustomed to the regime being imposed. Simultaneous measurements of the ECG and the respiration signals were then initiated, and continued for 180 sec. For each new measurement, the respiration frequency f_resp was increased by 1 breath per min ~i.e., by 1 min^-1). The minimal and maximal respiration frequencies were selected experimentally to be 3 and 30 min^-1, respectively, so that they should not be too difficult for the subject. Thus, 28 dataset pairs were measured for each subject. In between measurements the subject was allowed to relax for about 5 min until his heart rate, averaged over 1 min, had stopped changing. The average heart rate at rest was recorded to provide the current value of f₀before each new measurement. f₀will be used below to plot synchronization tongues.

Data processing
The phases for both measured signals were introduced using formula:
Below it is shown how time moments T_i were extracted from both types of data measured.

Regions of cardiorespiratory synchronization
The presence of n:m synchronization was arbitrarily defined by the condition that
In the Figures below the shaded areas correspond to parameters of respiration where the above condition for synchronization is satisfied during at least 5 successive periods of respiration and at least 20 sec.

Statistics
Effectiveness of the more common orders of synchronization n:m, estimated as an average duration of the plateaus of phase difference.		Average width of synchronization region.

More details can be found in

S. Rzeczinski, N. B. Janson, A. G. Balanov, and P. V. E. McClintock,"Regions of cardiorespiratory synchronization in humans under paced respiration", Phys. Rev. E. 66, 051909 (2002).

Please do not hesitate to contact Dr. Natalia Janson (n.b.janson@lboro.ac.uk) and Dr. Alexander Balanov (a.balanov@lboro.ac.uk) for reprints and discussions.