Sex differences in Actigraph‐estimated versus Self‐reported sleep quality and Alzheimer’s disease neuropathology among cognitively‐unimpaired older adults

Abstract

BackgroundDespite implications of sleep disturbances in Alzheimer’s disease (AD) pathophysiology (e.g. cerebral amyloid accumulation, atrophy in AD‐pathology sensitive brain regions); contradictory findings on sleep/neuropathology relationships at preclinical‐AD stage exist, possibly due to sex‐differences and/or methodological disparities and inconsistencies in sleep evaluation. We sought to address this conundrum using self‐reported versus actigraphy‐based data, to explore sex differences associated with amyloid load and MRI whole brain/sub‐regional volumetric indices, among asymptomatic older adults.MethodSex‐stratified analysis of CHARIOT:PRO SubStudy baseline data (519 cognitively‐unimpaired participants: 60–85 years) was performed, using a‐priori confounder‐adjusted multivariable regression models, aimed at assessing the effect‐modifying role of sex on associations of sleep quality measures (objective: 7‐day actigraphy, subjective: Pittsburg‐Sleep‐Quality‐Index questionnaire) with (i) Brain amyloid pathology status (positive/negative), (ii) volumetric‐MRI whole brain, (iii) hippocampal and (iv) frontal regions. Bivariate correlation analysis compared self‐reported and actigraphy sleep measures.Result49.1% of the participants were amyloid‐positive and more likely to report poor sleep. Only 5% of the cohort accurately estimated their overall sleep quality and were more likely to be younger (mean±SD: 69.1 ± 6.4), compared to inaccurate‐reporters (underestimated‐sleepQ: 72.7±5.5; overestimated‐sleepQ: mean 71.3±5.30; F (2, 336) = 5.18, P < 0.01). Women were more likely to underestimate their sleep quality compared to men (χ2 (2, 336) = 9.32, P < 0.01). Indeed, actigraphy revealed higher odds of insufficient sleep duration (OR, 1.75, 95% CI [1.13‐2.70]) and sleep efficiency impairments (OR, 4.36, 95% CI [1.47‐12.90] in men compared to women. Effect modification on associations of neurodegeneration indices and global sleep quality by sex revealed better self‐reported sleep quality specifically for women with larger hippocampal volumes (β:1.07e‐03;CI:2.48e‐04,1.89e‐03) and larger whole brain volumes (β:1.02e‐05;CI:3.06e‐06,1.74e‐05). Similarly, associations with amyloid pathology status showed worse actigraphy‐ascertained sleep efficiency for amyloid‐positive men (β:−1.44;CI:−2.79,−0.08).ConclusionActigraphy‐derived sleep quality metrices proved useful (compared to self‐reports) in differentiating sex dichotomies related to amyloid status and brain volumetric measures, emphasizing the importance of accounting for sex/gender divergence in subjective reporting, when exploring relationships between sleep quality and AD neuropathology. Further studies are warranted, extending upon these observations to explore potential predictive value for cognitive and functional trajectories from pre‐symptomatic stages.