Temporal Dynamics of Cerebral Blood Flow Reductions During Isocapnic Hyperoxia: The Role of Reactive Oxygen Species

Background: Hyperoxia is frequently used as medical treatment for, and prophylaxis to, hypoxia/ischemia-related conditions. In the brain, hyperoxia reduces cerebral blood flow (CBF), possibly through oxidative stress from reactive oxygen species (ROS) that disrupt vascular function; an effect that may intensify with prolonged exposure and could theoretically be mitigated by antioxidant (AO) treatment. However, the mechanisms underlying hyperoxia-induced CBF reductions remain poorly understood. We hypothesized that isocapnic hyperoxia would elicit a time-dependent decline in CBF at the gray matter (GM) and lobar levels, and that acute AO treatment would attenuate these reductions. Methods: Nineteen healthy adults (8F; 22 ± 5 yr; BMI 22 ± 2 kg/m 2 ) were studied; females were tested on days 1–5 of the menstrual cycle. In a randomized, double-blind, placebo-controlled design, participants ingested either a placebo (PLA) or an AO cocktail (α-lipoic acid, vitamin C, vitamin E) prior to MRI. CBF was measured using pseudo-continuous arterial spin labeling (pcASL) at 3T during normoxia and during isocapnic hyperoxia (~90% O 2 ) at ~5 (H1), 10 (H2), 25 (H3), and 30 min (H4). GM and six cerebral lobes were analyzed. End-tidal CO 2 , heart rate, and mean arterial pressure were continuously recorded. Results: GM CBF significantly decreased with hyperoxia, falling from 42 ± 12 mL/100 g/min at normoxia to 36 ± 11 (H1), 34 ± 10 (H2), 32 ± 10 (H3), and 33 ± 9 mL/100 g/min (H4) with placebo, and from 45 ± 13 to 38 ± 11, 34 ± 11, 35 ± 12, and 34 ± 12 mL/100 g/min with AO (all P < 0.05). In the brainstem/cerebellum, CBF declined by 17 ± 25% (H1) and 4 ± 36% (H4) with placebo, and by 11 ± 24% and 4 ± 32% with AO, although these changes were not significant (all P > 0.05). Subcortical CBF significantly decreased by 8 ± 10% and 10 ± 14% with PLA, and by 10 ± 10% and 21 ± 17% with AO (all P < 0.05). Frontal lobe CBF decreased by 10 ± 9% and 17 ± 11% with PLA, and by 12 ± 12% and 24 ± 14% with AO (all P < 0.05). Temporal lobe CBF fell by 11 ± 11% and 17 ± 13% with PLA, and by 13 ± 11% and 25 ± 14% with AO (all P < 0.05). Parietal lobe CBF decreased by 15 ± 9% and 18 ± 7% with PLA, and by 16 ± 13% and 27 ± 13% with AO (all P < 0.05). Occipital lobe CBF declined by 17 ± 14% and 20 ± 17% with PLA, and by 17 ± 17% and 24 ± 16% with AO (all P < 0.05). However, no significant differences were observed between placebo and AO across any timepoint at the GM or lobar level (all P > 0.05). Discussion: Contrary to our hypothesis, the magnitude of hyperoxia-induced CBF reductions did not differ between placebo and AO visits, either globally or across brain lobes, suggesting that ROS may not be the main driver of the observed cerebrovascular constriction. Clinically, this implies supplemental oxygen should be used with caution, as it could paradoxically limit oxygen delivery to the brain and increase the risk of neuronal injury. Future studies should investigate alternative mechanisms and examine longer exposures or vulnerable populations to improve clinical translation. This abstract was presented at the American Physiology Summit 2026 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.