Cerebrovascular Responses During Voluntary Breath Holding are Larger than Rebreathing

Abstract

Both voluntary breath-holding and rebreathing of expired air elicit changes in respiratory gas chemostimuli (CO2 and O2) at the metabolic rate. These chemostimuli elicit increases in cerebral blood flow (CBF), proportional to the magnitude of concomitant increases in CO2 and reductions in O2. These chemostimuli also activate the sympathetic nervous system, which increases systemic blood pressure. Although blood pressure responses appear small during rebreathing, evidence from obstructive sleep apnea (OSA) patients suggests higher mean blood pressure during sleep in those with worse OSA. We aimed to assess how superimposed changes in blood gases and increases in blood pressure affect the CBF responses during breath holding vs. rebreathing. We recruited 23 healthy participants (12 females) and instrumented them with a finometer (for beat-by-beat mean arterial blood pressure; MAP), transcranial Doppler ultrasound (middle and posterior cerebral artery velocity; MCAv, PCAv) and a pneumotachometer with gas sampling via a dual gas analyzer to assess the pressure of end-tidal (PET)CO2 and O2. Participants carried out two protocols in randomized order: (a) a maximal, voluntary end-inspiratory breath hold (BH) and (b) a rebreathing (RB) test. A breath-by-breath stimulus index (SI) was calculated as PETCO2/PETO2 during rebreathing, whereas the end-tidal gas values used to calculate SI were interpolated during breath holding from initial and break point values. During both BH and RB, cerebrovascular reactivity (CVR) was calculated as the MCAv or PCAv/SI. MAP reactivity (MAPR) was calculated as the slope of the MAP/SI response. Cerebrovascular conductance (CVC; MCA or PCA/MAP) reactivity (CVCR) was calculated as the slope of the MCACVC or PCACVC/SI responses. We found that (a) CVR was larger during BH vs. RB (MCA: 167.5±102.0 vs. 38.8±20.5 cm/s/SI, P<0.0001, n=23; PCA: 76.3±40.1 vs. 26.0±11.0 cm/s/SI, P<0.0001, n=19), (b) MAPR during BH was significantly higher than during RB (134.0±102.7 vs. 31.0±12.6 mmHg/SI, P=0.0001, n=23). and (c) CVCR during BH vs. RB (MCA: 0.75±0.69 vs. 0.16±0.15 cm/s/mmHg/SI, P<0.001, n=22; PCA: 0.26±0.29 vs. 0.12±0.09 cm/s/SI, cm/s/mmHg/SI, P=0.03, n=19). Our data demonstrate that breath holding elicited ~4-fold increases in MAP, translating to a larger anterior and posterior CVR compared to rebreathing. These findings suggest that the sympathetic responses during voluntary breath holding were larger than those during rebreathing across similar chemostimuli, potentially due to the differential sympathetic effects of struggling against a closed glottis during breath holding. This is the first evidence that voluntary apnea has larger effects on brain blood flow beyond that elicited by blood gas stimuli alone. Our data may have implications for understanding stroke risk in clinical populations with obstructive sleep apnea, whereby patients experience intermittent breath holds throughout the night, with associated spikes in arterial blood pressure.