Respiration-related cerebral blood flow variability increases during control-mode non-invasive ventilation in normovolemia and hypovolemia.

Abstract

Increased variability in cerebral blood flow (CBF) predisposes to adverse cerebrovascular events. Oscillations in arterial blood pressure and PaCO2 induce CBF variability. Less is known about how heart rate (HR) variability affects CBF. We experimentally reduced respiration-induced HR variability in healthy subjects, hypothesizing that CBF variability would increase. Internal carotid artery (ICA) blood velocity was recorded by Doppler ultrasound in ten healthy subjects during baseline, control-mode, non-invasive mechanical ventilation (NIV), i.e., with fixed respiratory rate, hypovolemia induced by lower body negative pressure, and combinations of these. ICA beat volume (ICABV) and ICA blood flow (ICABF) were calculated. HR, mean arterial blood pressure (MAP), respiratory frequency (RF), and end-tidal CO2 were recorded. Integrals of power spectra at each subject's RF±0.03Hz were used to measure variability. Phase angle/coherence measured coupling between cardiovascular variables. Control-mode NIV reduced HR variability (-56%, p=0.002) and ICABV variability (-64%, p=0.006) and increased ICABF variability (+140%, p=0.002) around RF. NIV+hypovolemia reduced variability in HR and ICABV by 70-80% (p=0.002) and doubled ICABF variability (p=0.03). MAP variability was unchanged in either condition. Respiration-induced HR and ICABV oscillations were in inverse phase and highly coherent (coherence >0.9) during baseline, but this coherence decreased during NIV, in normovolemia and hypovolemia (p=0.01). Controlling respiration in awake healthy humans reduced HR variability and increased CBF variability in hypovolemia and normovolemia. We suggest respiration-induced HR variability to be a mechanism in CBF regulation. Maintaining spontaneous respiration in patients receiving ventilatory support may be beneficial also for cerebral circulatory purposes.