Background: The brain is exceptionally reliant on a pervasive supply of oxygen, which is normally precisely matched to neural metabolic demand , via neurovascular coupling. Hypoxic exposure, (e.g. high altitude pursuits), challenges this tightly coupled relationship. As arterial oxygen saturations fall, global cerebral blood flow (gCBF) increases to maintain cerebral oxygen delivery, mitigating any threat of decoupling between oxygen supply and demand. However, we have recently shown some brain regions, including the posterior cingulate cortex (PCC) have an unexpected decrease in CBF during acute hypoxia ⁽¹⁾ suggesting a change in metabolism within these regions. Recent work investigating the cerebral metabolic rate of oxygen (CMRO₂) in hypoxia has shown a link to an increase in the concentration of the excitatory neurotransmitter glutamate ⁽²⁾. Changes in glutamate have also being linked to changes in neural activity and the hemodynamic response (neurovascular coupling) in task based studies ^(3,4) We hypothesised that if neurovascular coupling is unaffected by hypoxia, the regional reductions in CBF during hypoxia, suggesting reduced regional neural activity, should be reflected via an associated reduction in glutamate levels. Understanding the effects of hypoxia on neurovascular coupling and neurometabolism is important in understanding cognitive alterations that have been reported at altitude, and pathological conditions with a hypoxic component.
Methods: The present study exposed 11 participants to a moderate hypoxic environment (Fraction of inspired oxygen [FIO₂] = 0.12) for 3.5 hours and a procedurally matched normoxia condition (FIO₂ =.209). After 2 hours resting in an environmental chamber, participants were placed into a 3 T Philips Achieva MRI scanner whilst remaining in the hypoxic or normoxic condition, respectively. Whole brain resting state microvascular perfusion was quantified by pulsed Arterial Spin Labelling (22 slices, 2 x 2 x 5 mm in plane resolution, 1600 ms delay, TR 3 s, TE 15 ms, ), the regional resting state neurochemical environment was measured by PRESS ¹H Magnetic Resonance Spectroscopy (MRS), (TE = 40 ms, TR = 2 s, voxel size 20 x 20 x 20 mm³) within the PCC. Perfusion maps were calculated using the BASIL ⁽⁵⁾ software package and registered to MNI space for group comparison using FLIRT ⁽⁶⁾. MRS spectra were analysed using TARQUIN ⁽⁷⁾ 4.3.10, referencing to an unsuppressed water peak for metabolite quantification. All metabolite concentrations were corrected using partial volume correction for grey matter, white matter and CSF before comparison.
Results: Paired samples t-tests (cluster mass FWE correction at P<0.05) of the ASL data confirmed a reduction in perfusion during hypoxia compared to normoxia within the PCC and right posterior temporal cortex. In contrast MRS within the PCC revealed no significant change between normoxia M=15.76, SD=2.62 mM and hypoxia M=15.81, SD=3.91 mM in the excitatory neurotransmitter glutamate (p=0.9) nor any other major metabolite including creatine (p=0.7) , total n-acetyl aspartate (p=0.6) myo-inositol (p=0.7) and glutamine (p=0.3). However, a significant increase in Lactate was measured between normoxia M=0.57 SD=0.18 and hypoxia M=1.1, SD= 0.42, (p=0.01).
Conclusions: The present study supports our previous findings that indicate hypoxia induces a regional reduction in CBF within the PCC. Notably the PCC did not display a concomitant decrease in glutamate levels, however the measured increase in lactate would suggest an upregulation of non-oxidative metabolism. This indicates sustained neural activity and metabolic demand in presence of insufficient oxygen supply. Our data findings indicate this may be resulting from a hypoxia induced reduction in rCBF to the PCC without a concomitant reduction in neural activity. This challenges our present understanding of neurovascular coupling, whereby CBF is matched to demand, which is required to sustain healthy neural functioning. The findings suggest that hypoxia disrupts neurovascular coupling in a regionally specific manner, this may provide insight to specific cognitive deficits experienced at altitude.

Acknowledgements

Technical support and colleagues in the School of Psychology, Bangor University and the School of Sport Health and Exercise Sciences, Bangor University.