Sleep-disordered breathing may alter grey matter volume and glucose metabolism of the Default Mode Network in AD and MCI patients

1. BackgroundThe sleep-disordered breathing (SDB) describes a group of disorders characterized by abnormal respiratorypatterns or insufficient ventilation during sleep. It has been demonstrated that SDB is linked with an earlierage at cognitive decline onset with intermittent brain hypoxia and sleep fragmentation as possible underlyingpathophysiology. Neuroimaging studies have shown evidence of gray matter atrophy in subjects with SDB.However, the exact pathophysiology in the brain has not been investigated among subjects with concurrentmild cognitive impairment (MCI) and Alzheimers Disease (AD).2. MethodData were downloaded from Alzheimers Disease Neuroimaging Initiative (ADNI) database. The SDB groupingwas inferred from the self-reported medical history of subjects. Using 1:1 propensity score matching, SDB+and SDB- subjects were included considering several covariates: age, sex, education years, handedness,body mass index, use of continuous positive airway pressure (CPAP) or surgery as treatment, and subjectscognitive status (AD/MCI/NL), as well as protocol and field strength of T1 images. MR images with 1.5 & 3Tesla T1 MPRAGE/SPGR and 1.2 x 1.2 x 1.2 mm voxel size were preprocessed considering field distortion, andcorrection for subjects motion, normalized to MNI space and analyzed using Statistical Parametric Mapping(SPM) version 12 Voxel-based morphometry (VBM) to assess grey matter (GM) volume in the whole brain andwithin the DMN. Amygdala/hippocampus and insula masks were separately implemented to evaluate regionalGM volume, as previously proposed involved regions. Results were corrected for multiple comparisons usingthe FDR method.3. ResultsIn the whole brain and DMN masking analyses, total DMN size between SDB+ and SDB- subjects was notsignificantly different; regardless of subjects cognitive status. The DMN size was dramatically decreased indemented subjects compared to NL patients. Areas of DMN mostly being affected were bilateral amygdala,hippocampal and entorhinal cortices. In AD/MCI SDB+ vs. NL SDB+ comparison, both amygdala andhippocampus sizes were diffusely decreased. While, in AD/MCI SDB- patients compared with NL SDB- ones,only small regions of the left amygdala showed significant volume change. No significant difference could beextracted from any single cognitive condition with sleep-disordered breathing compared to its SDB-negativecounterparts, which is in contrast with previous findings on subjects without cognition deficits. This findingmay suggest an acceleration of amygdala atrophy in the context of SDB. 4. ConclusionsLack of difference in brain volume between SDB+ and SDB- subjects was in contrast with previous evidence.This may be due to the low sample size of the study, while the diminished effect of SDB in GM volumebetween subjects who had already diagnosed with AD and developed brain atrophy is possible. However,the difference in brain volume between MCI/AD and NL subjects were considerably higher in SDB+ subjects,especially in hippocampi and amygdala as regions involved in SDB pathophysiology. This may be indicative ofan underlying interaction or synergism between SDB and AD course in brain atrophy. Investigation of causalinference and the exact role of SDB in AD brain atrophy remains to be investigated in longitudinal designsstudies and larger sample sizes.