Background: Voluntary asphyxia imposed by static apnea represents a unique model to test the functional-structural limits of the neurovascular unit (NVU) in humans exposed to pathological extremes of hypoxaemia and hypercapnia. In the present study, we examined if [1] apnea would be associated with exaggerated cerebral oxidative-nitrosative stress (OXNOS) and structural destabilisation of the NVU and [3] distinguish between hypoxaemia and hypercapnea as the dominant vasoactive stimulus.

Methods: Ten ultra-elite breath-hold divers (6 male/4 female) aged 33 (mean) ± 9 (SD) years old performed two maximal dry apneas after normoxic hyperventilation (NX: severe hypoxaemia-hypercapnia) and hyperoxic hyperventilation (HX: absence of hypoxaemia while exacerbating hypercapnia) with measurements obtained at eupnea and after apnea. Mean arterial (MAP) and internal jugular venous (IJVP) pressures were recorded directly. The latter served as a validated surrogate of intracranial pressure (ICP). Net transcerebral biomarker exchange was calculated as the product of global cerebral blood flow (gCBF, duplex ultrasound) and radial arterial to internal jugular venous concentration gradients of plasma ascorbate free radical (A^•–, electron paramagnetic resonance spectroscopy), plasma nitrite (NO2-, ozone-based chemiluminescence) and select panel of serum NVU proteins (ELISA, Single Molecule Array). Following confirmation of distribution normality (Shapiro-Wilk W tests), data were analysed using a combination of two (State: eupnea vs. apnea × Site: arterial vs. venous) and three (Trial: NX vs. HX × State × Site) factor repeated measures ANOVAs with post-hoc Bonferroni-corrected paired samples t-tests.

Results: Compared to HX, greater increases in MAP (+61 ± 9 vs. +47 ± 14 mmHg, P = 0.021) and IJVP (+12 ± 2 vs. +7 ± 5 mmHg, P = 0.005) were observed during apnea in NX whereas the increase in gCBF was lower (+83 ± 22 % vs. +206 ± 52 %, P = <0.001). Apnea in NX stimulated a greater net cerebral output (venous > arterial) of A^•–  (-4049 ± 5035 vs. -609 ± 5273 AU/100g/min, P = 0.042) and lower uptake (arterial > venous) of NO2- (1787 ± 2029 vs. 3806 ± 2334 nM/100g/min, P = 0.036), highlighting the key contribution of hypoxaemia to OXNOS. This coincided with a greater net cerebral output of S100B, glial fibrillary acidic protein, ubiquitin carboxy-terminal hydrolase L1, neurofilament light-chain and total tau (all P < 0.05).

Conclusions: Collectively, these findings demonstrate that NVU integrity is more impaired during extreme apnea-induced hypoxemic- compared to hyperoxemic-hypercapnic stress highlighting hypoxia as a key stimulus underlying a transient increase in blood-brain barrier permeability and neuro-gliovascular reactivity/damage. Structural changes were linked to the combined elevation in molecular (↑OXNOS) and haemodynamic (↑systemic/intracranial hypertension) stress. Collectively, these novel findings provide a potential mechanism whereby the combined effects of molecular-haemodynamic stress to which the ‘apneic brain’ is exposed converge at the NVU transiently compromising integrity and function.