Introduction

The brain is reliant on a supply of oxygen to meet neural metabolic demand via a relationship known as neurovascular coupling. Exposure to hypoxia such as that at high-altitude challenges this tightly coupled relationship. As arterial oxygen saturation falls, cerebral blood flow (CBF) increases to maintain oxygen delivery to mitigate any threat of decoupling between oxygen supply and demand. We have shown that the posterior cingulate cortex (PCC) has an unexpected decrease in CBF during acute hypoxia^(1,2). Furthermore, during a memory recall task, hypoxia reversed the task evoked BOLD signal within the PCC, despite no change in performance suggesting neural activity was unchanged ⁽²⁾. These findings suggest that hypoxia causes a regionally-specific reversal of neurovascular coupling; however, the inherent vascular dependence of the BOLD signal means it cannot distinguish between vascular and metabolic alterations that may underlie this observation. This study utilised functional MRS (fMRS) to measure the metabolite dynamics within the PCC during a memory recall task during hypoxia. It was hypothesised that glutamate concentration within the PCC will increase in response to a memory recall task during normoxia, and that hypoxia will not change this task-induced glutamate response.

Method

In a repeated-measures crossover design 15 participants were exposed to 3.5 hours of normoxia (fraction of inspired oxygen; FiO₂=20.9%) and a procedurally matched poikilocapnic hypoxia (FiO₂=12.0%). After 2 hours resting, participants were scanned at 3T (Philips, Achieva) whilst remaining in the respective condition. Event-related fMRS (PRESS, TE=40ms, TR=2s, voxel size 20x20x20 mm³) allowed the probing of the dynamic alterations in metabolites during rest and memory recall in both conditions. Metabolite concentrations were estimated using QUEST (jMRUI), referencing to an unsuppressed water peak for absolute concentration estimations. A 2x2 repeated measures ANOVA was used to assess differences in metabolite concentrations between rest and task in both conditions.

Results

Glutamate increased by 9.1% between rest and memory recall in normoxia (95%CI: [16, 2]; P=0.01; Figure 1). This task-induced increase was not present in hypoxia with participants on average showing a 3.5% reduction (95%CI: [CI 2, -9]; P>0.05; Figure 1). As a result, comparing across condition glutamate was observed to be 15.9% (Mm) lower during memory recall in hypoxia (95%CI: [-7, -25]; P=0.001) however condition-specific rest did not significantly differ. Further exploratory analyses revealed a significant main effect of condition on glucose concentration. Glucose concentration in hypoxia was lower during rest (0.70±0.20 vs 0.57±0.20, P=0.01) and recall (0.72±0.21 vs 0.56±0.23, P=0.01). No other measured metabolite was found to change.

Discussion

Loss of task-induced increases in glutamate during hypoxia suggest oxidative metabolism is altered within the PCC. Glutamate has been demonstrated as a modulator of NVC, thus disruption to its normal task induced changes could lead to an altered vascular response in the PCC during task. However, the reduction in glucose seen during hypoxia would suggest the PCC is still metabolically active, but this may be sustained by non-oxidative means.

Acknowledgements

Technical and support staff based within the School of Sport Health and Exercise Sciences and School of Psychology, Bangor University.