Subclinical depressive symptoms are associated with increased risk of Alzheimer’s disease (AD), but the brain mechanisms underlying this relationship are still unclear. We aimed to provide a comprehensive overview of the brain substrates of subclinical depressive symptoms in cognitively unimpaired older adults using complementary multimodal neuroimaging data. We included cognitively unimpaired older adults from the baseline data of the primary cohort Age-Well (n = 135), and from the replication cohort ADNI (n = 252). In both cohorts, subclinical depressive symptoms were assessed using the 15-item version of the Geriatric Depression Scale; based on this scale, participants were classified as having depressive symptoms (>0) or not (0). Voxel-wise between-group comparisons were performed to highlight differences in gray matter volume, glucose metabolism and amyloid deposition; as well as white matter integrity (only available in Age-Well). Age-Well participants with subclinical depressive symptoms had lower gray matter volume in the hippocampus and lower white matter integrity in the fornix and the posterior parts of the cingulum and corpus callosum, compared to participants without symptoms. Hippocampal atrophy was recovered in ADNI, where participants with subclinical depressive symptoms also showed glucose hypometabolism in the hippocampus, amygdala, precuneus/posterior cingulate cortex, medial and dorsolateral prefrontal cortex, insula, and temporoparietal cortex. Subclinical depressive symptoms were not associated with brain amyloid deposition in either cohort. Subclinical depressive symptoms in ageing are linked with neurodegeneration biomarkers in the frontolimbic network including brain areas particularly sensitive to AD. The relationship between depressive symptoms and AD may be partly underpinned by neurodegeneration in common brain regions.

Study objectives:To determine the polysomnography characteristics during sleep paralysis, false awakenings and lucid dreaming (which are states intermediate to REM sleep and wake but exceptionally observed in sleep laboratory). Methods:In 5 subjects, we captured 5 episodes of sleep paralysis (2 time-marked with the ocular left-right-left-right code normally used to signal lucid dreaming, 1 time-marked by an external noise and 2 retrospectively reported) and 2 episodes of false awakening. The sleep coding (using three seconds mini-epochs) and spectral EEG analysis were compared during these episodes and normal REM sleep as well as wakefulness in the same four among these five subjects, and vs. lucid REM sleep in four other patients with narcolepsy.Results: During episodes of sleep paralysis, 70.8 % of mini-epochs contained theta EEG rhythm, (vs. 89.7% in REM sleep and 21.2% in wakefulness), 93.8% contained chin muscle atonia (vs. 89.7% in REM sleep and 33.3% in wakefulness) and 6.9% contained rapid eye movements (vs. 11.9% in REM sleep and 8.1% in wakefulness). The EEG spectrum during sleep paralysis was intermediate between wakefulness and REM sleep in the alpha, theta and delta frequencies, whereas the beta frequencies were not different between sleep paralysis and normal REM sleep.The power spectrum during false awakening followed the same profile as in sleep paralysis. Conclusions:The predominant theta EEG rhythm during sleep paralysis and false awakenings (with rare and lower alpha rhythm) suggests that the brain during sleep paralysis is not in an awake but in a dreaming state. 4 BRIEF SUMMARYCurrent Knowledge/Study Rationale: Sleep paralysis is commonly viewed as an awake brain with a complete muscle atonia, but EEG analysis during sleep paralysis is exceptional.Study Impact: Our polysomnography and spectral analysis of time-marked sleep paralyses suggest that the brain during sleep paralysis is not in an awake but in a dreaming state. This may explain why hallucinations are often associated with sleep paralysis.

Sleep plays a crucial role in memory consolidation. Recent data in rodents and young adults revealed that fast spindle band power fluctuates at a 0.02-Hz infraslow scale during non-rapid eye movement (NREM) sleep. These fluctuations result from a periodic temporal clustering of spindles and may modulate sleep maintenance and memory consolidation. With age, sleep undergoes substantial changes but age-related changes in spindle clustering have never been investigated. Polysomnography data were collected in 147 older (mean age ± SD: 69.3 ± 4.1 years) and 32 young-middle aged (34.5 ± 10.9 years) adults. Sleep-dependent memory consolidation was assessed in a subsample of 57 older adults using a visuospatial memory task. We analyzed power fluctuations in fast spindle frequency band, detected fast spindles and quantified their clustering during the night separating encoding and retrieval. Fast spindle band power fluctuated at a 0.02-Hz infraslow scale in young-middle aged and older adults. However, the proportion of clustered fast spindles decreased non-linearly with age (p < 0.001). This effect was not mediated by NREM sleep fragmentation. The clustering level of fast spindles modulated their characteristics (p < 0.001). Finally, the mean size of spindle clusters was positively associated with memory consolidation (p = 0.036) and negatively with NREM sleep micro-arousals density (p = 0.033). These results suggest that clusters of fast spindles may constitute stable sleep periods promoting off-line processes such as memory consolidation. We emphasize the relevance of considering spindle dynamics, obviously impaired during ageing, to understand the impact of age-related sleep changes on memory.

Objectives: Deepening our understanding of the mechanisms by which meditation practices impact well-being and human flourishing is essential for advancing the science of meditation.The phenomenologically grounded classification system introduced by Dahl, Lutz, and Davidson (2015) distinguishes attentional, constructive, and deconstructive forms of meditation based on the psychological mechanisms these practices primarily target or necessitate. Our main aim was to understand whether this theory-based taxonomy could be used as a guiding principle for combining established psychological self-report measures of meditation-related mechanisms into psychometrically adequate composite scores. Methods:We used cross-sectional data to compute meditation composite scores in three independent samples, namely meditation-naïve healthy older adults from the Age-Well trial (n = 135), meditation-naïve older adults with subjective cognitive decline from the SCD-Well trial (n = 147), and healthy long-term meditators (≥10,000 hours of practice including one three-year meditation retreat) from the Brain & Mindfulness project (n=29). The psychometric properties of the composite scores were assessed via floor and ceiling effects, composite intercorrelations, interpretability, and convergent validity in relation to well-being, anxiety, and depression.Results: Three theoretically derived meditation composite scores, reflecting mechanisms involved in attentional, constructive, and deconstructive practices, displayed adequate psychometric properties. Separate secondary confirmatory factor analyses empirically corroborated the theoretically predicted three-factor structure of this classification system. Conclusions:Complementing data-driven approaches, this study offers preliminary support for using a theoretical model of meditation-related mechanisms to create empirically meaningful and psychometrically sound composite scores. We conclude by suggesting conceptual and methodological considerations for future research in this area.

Rapid eye movement (REM) sleep is markedly altered in Alzheimer's disease (AD), and its reduction in older populations is associated with AD risk. However, little is known about the underlying brain mechanisms. Our objective was to investigate the relationships between REM sleep integrity and amyloid deposition, gray matter volume, and perfusion in aging. Methods: We included 121 cognitively unimpaired older adults (76 women, mean age 68.96 AE 3.82 years), who underwent a polysomnography, T1-weighted magnetic resonance imaging, early and late Florbetapir positron emission tomography scans to evaluate gray matter volume, perfusion, and amyloid deposition. We computed indices reflecting REM sleep macro-and microstructural integrity (ie, normalized electroencephalographic spectral power values). Voxelwise multiple regression analyses were conducted between REM sleep indices and neuroimaging data, controlling for age, sex, education, the apnea-hypopnea index, and the apolipoprotein E ε4 status. Results: Lower perfusion in frontal, anterior and posterior cingulate, and precuneus areas was associated with decreased delta power and electroencephalographic slowing (slow/fast frequencies ratio), and increased alpha and beta power. To a lower extent, similar results were obtained between gray matter volume and delta, alpha, and beta power. In addition, lower REM sleep theta power was more marginally associated with greater diffuse amyloid deposition and lower gray matter volume in fronto-temporal and parieto-occipital areas. Interpretation: These results suggest that alterations of REM sleep microstructure are associated with greater neurodegeneration and neocortical amyloid deposition in older adults. Further studies are warranted to replicate these findings, and determine whether older adults exhibiting REM sleep alterations are more at risk of cognitive decline and belonging to the Alzheimer's continuum.

Background This study assesses the relationships between dynamic functional network connectivity (DFNC) and dementia risk. Methods DFNC of the default mode (DMN), salience (SN), and executive control networks was assessed in 127 cognitively unimpaired older adults. Stepwise regressions were performed with dementia risk and protective factors and biomarkers as predictors of DFNC. Results Associations were found between times spent in (i) a “weakly connected” state and lower self-reported engagement in early- and mid-life cognitive activity and higher LDL cholesterol; (ii) a “SN-negatively connected” state and higher blood pressure, higher depression score, and lower body mass index (BMI); (iii) a “strongly connected” state and higher self-reported engagement in early-life cognitive activity, Preclinical Alzheimer’s cognitive composite-5 score, and BMI; and (iv) a “DMN-negatively connected” state and higher self-reported engagement in early- and mid-life stimulating activities and lower LDL cholesterol and blood pressure. The lower number of state transitions was associated with lower brain perfusion. Conclusion DFNC states are differentially associated with dementia risk and could underlie reserve.

Background and objectives:Sleep-disordered breathing (SDB) has been related to amyloid deposition and increased dementia risk. However, how SDB relates to medial temporal lobe neurodegeneration and subsequent episodic memory impairment is unclear. Our objective was to investigate the impact of amyloid positivity on the associations between SDB severity, medial temporal lobe subregions and episodic memory performance in cognitively unimpaired older adults.Methods:Data were acquired between 2016 and 2020 in the context of the Age-Well randomized controlled trial of the Medit-Ageing European project. Participants aged over 65 years old, free of neurological, psychiatric or chronic medical diseases were recruited from the community. They completed a neuropsychological evaluation, in-home polysomnography, a Florbetapir positron emission tomography and magnetic resonance imaging, including a specific high-resolution assessment of the medial temporal lobe and hippocampal subfields. Multiple linear regressions were conducted to test interactions between amyloid status and SDB severity on the volume of MTL subregions, controlling for age, sex, education and the ApoE4 status. Secondary analyses aimed at investigating the links between SDB, MTL subregional atrophy and episodic memory performance at baseline and at a mean follow up of 20.66 months in the whole cohort and in subgroups stratified according to amyloid status.Results:We included 122 cognitively intact community-dwelling older adults (mean age ± SD: 69.40 ± 3.85 years, 77 women, 26 Aβ+ individuals) in baseline analyses and 111 at follow-up. The apnea-hypopnea index interacted with entorhinal (β=-0.81, p<0.001, pη2=0.19), whole hippocampal (β=-0.61, p<0.001, pη2=0.10), subiculum (β=-0.56, p=0.002, pη2=0.08), CA1 (β=-0.55, p=0.002, pη2=0.08) and DG (β=-0.53, p=0.003, pη2=0.08) volumes, such that higher sleep apnea severity was related to lower MTL subregions volumes in amyloid-positive individuals, but not those who were amyloid-negative. In the whole cohort, lower whole hippocampal (r=0.27, p=0.005) and CA1 (r=0.28, p=0.003) volumes at baseline were associated with worse episodic memory performance at follow-up.Discussion:Overall, we showed that SDB was associated with MTL atrophy in cognitively asymptomatic older adults engaged in the Alzheimer’s continuum, which may increase the risk of developing memory impairment over time.Trial Registration Information:ClinicalTrials.gov Identifier:NCT02977819

Background Sleep‐dependent memory consolidation, which is thought to rely on the dialogue between the hippocampus and the medial prefrontal cortex (mPFC) during NREM sleep, is mediated by slow waves (SW) and sleep spindles. Some studies indicate that this process is impaired in ageing but also in sleep‐disordered breathing (SDB), a very common sleep disorder in ageing. In this study, we investigated sleep‐dependent memory consolidation in older adults with or without SDB and explored the underlying mechanisms. Method Baseline data of 57 cognitively unimpaired older adults (mean age ± SD: 68.6 ± 3.3 years) from the Age‐Well cohort were analysed. Participants underwent a structural MRI scan to obtain mPFC and hippocampal grey matter volumes, and performed a visuospatial memory task until reaching a learning criterion of 66.6% of correct answers. Post‐learning sleep was monitored using polysomnography, and delayed recall was probed the next morning. An overnight change in memory performance (OCMP) was computed as follows: (recall performance – learning performance) / learning performance. SW, slow and fast spindles during N2 and N3 sleep were automatically detected using the SleepTrip toolbox. Based on the standard apnea‐hypopnea index (AHI) cutoff of 15 events/h, participants were classified as having SDB (SDB+, n=47) or not (SDB‐, n=10). Result There was no between‐group difference in OCMP, as well as in amplitude, duration and frequency of SW, slow and fast spindles (p>0.05). However, we found a significant group by SW density interaction after controlling for age, sex, education, total sleep time and trait‐anxiety (p=0.03). Thus, a positive association was only found between SW density and OCMP in the SDB‐ group (p=0.02; Fig 1). No correlation was found between slow or fast spindle density and OCMP in either group. Finally, mPFC and hippocampal volumes did not differ between SDB+ and SDB‐ participants (p>0.05), but SW density positively correlated with the volume of the mPFC (p=0.01; Fig 2). Conclusion Our findings suggest that the contribution of SW to sleep‐dependent memory consolidation is impaired in participants with SDB and incriminate the mPFC. We plan to conduct connectivity and SW‐spindle coupling analyses to unravel the functional substrates of this deficit.