Sleep patterns differ across Alzheimer's disease phenotypes: Implications for selective vulnerability and customized treatment.

Abstract

Sleep-wake disturbances are a prominent feature in Alzheimer's Disease (AD), negatively impacting patients and caregivers' quality of life. Growing evidence suggests that sleep-wake dysregulation happens early in AD due to the AD-tau degeneration of the arousal system (i.e., locus coeruleus). However, whether the pattern of sleep-wake dysfunction is different among AD clinical variants and how it is different remains unclear. It is an unmet gap of high importance because these differing patterns may require tailored treatment strategies. Since AD phenotypes with different clinical profiles have distinct regional patterns of cortical tau, we hypothesized that AD variants may also have differential sleep patterns. Therefore, we aim to contrast the behavioral sleep features within AD variants. A cohort of 24 patients, including 8 amnestic AD and 16 non-amnestic AD (7 lvPPA, 9 PCA), with a neuropathological or biomarker based diagnosis of AD, underwent an overnight EEG monitoring. Sleep stages were evaluated by PRANA software and confirmed by REM Logic spectral analyses. We performed nonparametric analyses to study group differences in the amount of each sleep stage as a percent of total sleep time. Additionally, we performed linear regression models adjusting by the age of onset, MMSE, and CDR Total. There were no differences between amnestic AD, lvPPA, and PCA groups in gender, age of onset (59±11, 55±6, 59±8, respectively), global cognition (MMSE 21±7, 23±3, 21±5), or functional performance (CDR Total 0.9±0.5, 0.7±0.3, 1±0.4). Compared to non-amnestic AD, amnestic AD had more REM sleep (p<0.05) and a trend toward lower NREM sleep stages 1 and 3. In the same line, this pattern was still observed when comparing the three clinical variants (amnestic AD, lvPPA, and PCA) (p<0.05). Regression models confirmed that clinical AD variants accounted for differences in REM sleep (p<0.01). We are continuing to evaluate additional AD participants and age-matched healthy controls. Preliminary results suggest that sleep profiles might differ between AD variants. The existence of specific sleep-wake profiles would suggest differential degeneration patterns of the arousal system within AD phenotypes, which would open the door to uncover the selective vulnerability of subcortical structures for the first time.