The human electroencephalogram (EEG) of sleep undergoes profound changes with age. These changes can be conceptualized as "brain age", which can be compared to an age norm to reflect the deviation from normal aging process. Here, we develop an interpretable machine learning model to predict brain age based on two large sleep EEG datasets: the Massachusetts General Hospital sleep lab dataset (MGH, N = 2,621) covering age 18 to 80; and the Sleep Hearth Health Study (SHHS, N = 3,520) covering age 40 to 80. The model obtains a mean absolute deviation of 8.1 years between brain age and chronological age in the healthy participants in the MGH dataset. As validation, we analyze a subset of SHHS containing longitudinal EEGs 5 years apart, which shows a 5.5 years difference in brain age. Participants with neurological and psychiatric diseases, as well as diabetes and hypertension medications show an older brain age compared to chronological age. The findings raise the prospect of using sleep EEG as a biomarker for healthy brain aging.In total, we identify 2,621 EEGs where 189 of them have neurological or psychiatric diseases. Table 1 provides summary statistics for the dataset.

IMPORTANCE Aging is associated with excessive daytime sleepiness (EDS), which has been linked to cognitive decline in the elderly. However, whether EDS is associated with the pathologic processes of Alzheimer disease remains unclear.OBJECTIVE To investigate whether EDS at baseline is associated with a longitudinal increase in regional β-amyloid (Aβ) accumulation in a cohort of elderly individuals without dementia. DESIGN, SETTING, AND PARTICIPANTSThis prospective analysis included participants enrolled in the Mayo Clinic Study of Aging, a longitudinal population-based study in Olmsted County, Minnesota. Of 2900 participants, 2172 (74.9%) agreed to undergo carbon 11-labeled Pittsburgh compound B positron emission tomography (PiB-PET). We included 283 participants 70 years or older without dementia who completed surveys assessing sleepiness at baseline and had at least 2 consecutive PiB-PET scans from January 1, 2009, through July 31, 2016, after excluding 45 (13.7%) who had a comorbid neurologic disorder. MAIN OUTCOMES AND MEASURESExcessive daytime sleepiness was defined as an Epworth Sleepiness Scale score of at least 10. The difference in Aβ levels between the 2 consecutive scans (ΔPiB) in Aβ-susceptible regions (prefrontal, anterior cingulate, posterior cingulate-precuneus, and parietal) was determined. Multiple linear regression models were fit to explore associations between baseline EDS and ΔPiB while adjusting for baseline age, sex, presence of the apolipoprotein E ε4 allele, educational level, baseline PiB uptake, global PiB positivity (standardized uptake value ratio Ն1.4), physical activity, cardiovascular comorbidities (obesity, hypertension, hyperlipidemia, and diabetes), reduced sleep duration, respiratory symptoms during sleep, depression, and interval between scans. RESULTSOf the initial 283 participants, mean (SD) age was 77.1 (4.8) years; 204 (72.1%) were men and 79 (27.9%) were women. Sixty-three participants (22.3%) had EDS. Baseline EDS was significantly associated with increased regional Aβ accumulation in the anterior cingulate

Excessive daytime sleepiness (EDS) and fatigue increases with age. The aim of this study was to investigate the association between EDS and fatigue with cortical thickness and hippocampal volume in cognitively normal, late middle-aged and older adults. We performed a cross-sectional observational study of 1374 cognitively-normal subjects aged 50 years and older who had a structural MRI. Regional cortical thickness and hippocampal volume were measured. Multiple linear regression models were fit to explore associations between EDS and fatigue and structural MRI measures in different brain regions, adjusting for multiple covariates. EDS was defined as Epworth Sleepiness Scale ≥10. Fatigue severity was assessed with the Beck Depression Inventory-2. 208 participants had EDS, 27 had significant fatigue, and 11 had both. Participants with EDS or fatigue had significantly lower cognitive scores, more disturbed sleep, and medical comorbidities. The presence of EDS was associated with both global and regional atrophy, whereas fatigue was more associated with frontal and temporal changes. Cortical thinning predicted by EDS and fatigue was maximal in the temporal region with average reduction of 34.2 µm (95% CI, −54.1, −14.3; P=.001) and 90.2 µm (95% CI, −142.1, −38.2; P=.001), respectively. Fatigue was also associated with hippocampal volume reduction of −374.2 mm3 (95%CI, −670.8, −77.7; P=.013). Temporal cortical thinning predicted by presence of EDS and fatigue was equivalent to more than 3.5 and nine additional years of aging, respectively. EDS and fatigue were associated with cortical thickness reduction primarily in regions with increased age-susceptibility, which may indicate accelerated brain aging.

No abstract