Youquan Ning, Yilan Li, Zhenliang Xiong, Zehu Sheng, Lian He, Shaoxin Xiang, Yida Hu, Juan Peng, Fajin Lv
Our findings suggest that OSA severity is associated with ACC-specific BOLD-CSF decoupling and a disrupted overnight decline in ACC-GABA. The magnitude of overnight E/I variation moderated the association between BOLD-CSF coupling variation and cognitive performance. These preliminary findings highlight the potential relevance of ACC-BOLD-CSF coupling and neurochemical alterations to OSA-related cognitive impairment. Future studies with larger sample sizes are needed to confirm these findings.
OBJECTIVE: The roles of cortical glymphatic dysfunction and excitation-inhibition imbalance in cognitive impairment associated with obstructive sleep apnea (OSA) require further study. This study aimed to investigate sleep-related glymphatic function and neurotransmitter changes in OSA and their relationships with cognition.
METHODS: Eighty seven participants (50 OSA, 37 controls) completed a pre-sleep MRI, followed by an overnight PSG recording, and a post-sleep MRI to assess sleep-dependent glymphatic and neurotransmitter changes. Resting-state blood-oxygen-level-dependent (BOLD) signals and cerebrospinal fluid (CSF) signals (BOLD-CSF coupling), an indirect glymphatic-related imaging proxy, were assessed in the global cortex, anterior cingulate cortex (ACC), posterior cingulate cortex (PCC), and default mode network (DMN) cortex. MRS-derived gamma-aminobutyric acid (GABA) and glutamate/glutamine (Glx) levels were measured in the ACC and PCC as indirect neurochemical measures of excitation-inhibition (E/I) balance. General linear models, linear mixed-effects models, correlation analyses, and moderation analyses were employed to examine group differences, overnight changes in BOLD-CSF coupling and neurotransmitters, and their relationships with cognitive performance.
RESULTS: In the moderate-to-severe OSA group, pre-sleep global and ACC-BOLD-CSF coupling were weaker than in controls, and ACC-GABA showed a group × time interaction (ηp 2 = 0.105, p-FDR = 0.049), with a significant overnight decrease observed only in controls (Cohen's d = 0.75, p-FDR = 0.007). Greater magnitudes of overnight change in both ACC-BOLD-CSF coupling and ACC-E/I were independently associated with worse cognition. ACC-E/I change also moderated the coupling-cognition relationship (β = 0.590, p = 0.015), with a significant negative association only at smaller magnitudes of E/I change (β = -0.729, p = 0.002).
CONCLUSION: Our findings suggest that OSA severity is associated with ACC-specific BOLD-CSF decoupling and a disrupted overnight decline in ACC-GABA. The magnitude of overnight E/I variation moderated the association between BOLD-CSF coupling variation and cognitive performance. These preliminary findings highlight the potential relevance of ACC-BOLD-CSF coupling and neurochemical alterations to OSA-related cognitive impairment. Future studies with larger sample sizes are needed to confirm these findings.