Abstract
BACKGROUND: Older adults with type 2 diabetes (T2D) and mild cognitive impairment (MCI) are at increased risk for dementia. Hyperbaric oxygen therapy (HBOT) has shown vascular and metabolic effects that may benefit brain health. This randomized, double‐blind, sham‐controlled trial assessed the impact of HBOT on cognitive function, cerebral blood flow (CBF), and glucose metabolism in this high‐risk group. METHODS: A total of 155 participants were recruited and randomized to HBOT (N =77) and sham (N =78) between 2017 and 2023. Cognitive outcomes were assessed at baseline, 3, 6, and 12 months. Imaging (ASL‐MRI for CBF and FDG‐PET for glucose metabolism) was conducted at baseline, 3, and 12 months. The primary outcomes included global cognition (composite z‐score), CBF, and FDG‐PET. SUVR values were calculated using a global cortical ROI as the target region, normalized to the pons as the reference region. Secondary cognitive measures included domain‐specific cognition (executive functions and episodic memory) and CDR‐SOB. Analyses used linear mixed‐effects models under intent‐to‐treat (ITT) and per‐protocol (PP) approaches, without adjustment for multiple comparisons. RESULTS: Across all timepoints, participants improved in cognition compared to baseline. At 3 months, global cognition and executive function significantly favored sham (p <0.05), with no sustained group differences at later timepoints (see Figure). Memory improved over time in both groups, with no significant group differences. There were no differences in CBF or FDG‐PET SUVR between groups. Sex and baseline cognition (median CDR‐SOB) had no significant effects on treatment response. PP analyses suggested lower SUVR at 3 months in the HBOT group across several brain regions (p ‐values: 0.014‐0.048). No group differences were seen in correlations between changes in cognition and CBF or SUVR. Adverse event analysis showed three times more serious adverse events (SAEs) in the HBOT group (N =25) comparing to sham (N =8), distributed across multiple organ systems. AEs distribution was similar between groups. CONCLUSIONS: HBOT did not improve cognitive or brain imaging outcomes compared to sham in older adults with T2D and MCI. A transient cognitive benefit was observed in the sham group immediately post‐intervention. The higher rate of SAEs in the HBOT group for this population with high comorbidities warrants further investigation.