Introduction: 

Coupling between respiration and the heart rhythm at rest is known as respiratory sinus arrhythmia. Heart rate increases during inspiration and lowers during expiration. Independent of this modulation of heart rate, intrathoracic pressure changes during respiration cause periodic movements of the chest and this causes R peak amplitude changes in the ECG. This can be used to derive respiration from the ECG itself. We use ECG derived respiration and the ECG to investigate the coupling between respiration and the heart during sleep in order to look for sleep stage dependent changes in a large set of healthy subjects.

Methods: 

Cardiorespiratory polysomnography with a chest wall ECG and three channels for respiration (oronasal airflow, chest and abdominal movements) were obtained in 112 healthy subjects (SIESTA study). We derived in addition the respiratory signal from the ECG using the modulation of R wave amplitudes. The reliability of this trace was checked using cardiorespiratory polysomnography. With this we calculate the synchronisation of the respiratory waveform and the heart beat by calculating the instantaneous phase shift using the Hilbert transformation.

Results: 

The reconstruction of breathing from the ECG was reliable in most subjects. Based on this we reconstructed the breathing patterns from the ECG without additional signals. In healthy subjects we found 3.8% of the time in non-REM with coupled respiration and heart beat and only 0.6% of the time spent in REM sleep. During wakefulness within the sleep recording we found 1.6% of the time with coupled respiration and heart beat. This amount of time decreases with increasing age and decreases with body mass index. There is no gender difference.

Conclusion: 

It is possible to derive respiration from a combination of R-R interval analysis and R peak amplitude variability. The varying synchronization between the heart beat and respiration changes with sleep stage and can give indications for the severity of sleep related breathing disorders because it relates to the cardiorespiratory regulation responsible for pathophysiology.

Acknowledgement: This work had been supported by the European Union funded project DAPHNET (2006–2009).