Takumi Shiraishi, Toru Murakawa-Hirachi, Kojiro Nogami, Masahiro Ohgidani, Masahiro Natsuaki, Masashi Nishihara, Yutaka Kunitake, Hiroshi Tateishi, Jun Matsushima, Hiroko Kunitake, Ryuzo Orihashi, Ryohei Kojima, Junko Oishi, Jun Kikuchi, Chika Nagahama, Ken Takada, Masataka Hirano, Shota Shiba, Airi Fukai, Akira Tomonari, Daichi Inayoshi, Koichi Node, Akira Monji, Yoshito Mizoguchi
CPAP-induced brain recovery was characterized by coordinated, tissue-specific diffusion changes after 3 months of treatment. The combination of DTI and NODDI identified orientation-dispersion changes that are not captured by conventional DTI, while the absence of RD and NDI changes suggests that early recovery may precede detectable remyelination or neurite-density changes. These findings support NODDI-based tissue-compartment imaging as a promising approach for investigating the biological basis of brain recovery following CPAP in OSA.
OBJECTIVES: Continuous positive airway pressure (CPAP) can partially reverse brain abnormalities associated with obstructive sleep apnea (OSA), but the biological substrates underlying this recovery remain unclear. We investigated whether diffusion MRI could identify tissue-compartment changes associated with brain recovery following CPAP.
METHODS: Thirteen adults with OSA underwent paired brain MRI before and after 3 months of CPAP treatment. Diffusion tensor imaging (DTI) and neurite orientation dispersion and density imaging (NODDI) were analyzed using tract-based spatial statistics (TBSS; n = 13) and gray matter-based spatial statistics (GBSS; n = 10). Exploratory region-of-interest (ROI) analyses were performed across 102 Gy matter regions. Voxel-wise analyses used permutation-based inference with threshold-free cluster enhancement and family-wise error (FWE) correction.
RESULTS: CPAP produced three coordinated white matter changes that survived FWE correction: FA increased in 7 of 10 examined tracts, restricted to the left hemisphere; AD increased bilaterally across all 10 tracts; and ODI decreased bilaterally across all 10 tracts. No significant changes were detected in RD, MD, NDI, or FWF within white matter. In gray matter, FWE-corrected analysis identified increased FA in the right cerebellar Crus II and decreased FWF in the left medial orbitofrontal cortex. Exploratory ROI analysis suggested widespread cortical microstructural changes and distinct diffusivity patterns in subcortical regions, with changes concentrated within default mode and salience networks. AHI was significantly reduced following CPAP, and HAMD scores improved.
CONCLUSIONS: CPAP-induced brain recovery was characterized by coordinated, tissue-specific diffusion changes after 3 months of treatment. The combination of DTI and NODDI identified orientation-dispersion changes that are not captured by conventional DTI, while the absence of RD and NDI changes suggests that early recovery may precede detectable remyelination or neurite-density changes. These findings support NODDI-based tissue-compartment imaging as a promising approach for investigating the biological basis of brain recovery following CPAP in OSA.