Résumé
Objective: To determine the relationships between self-reported sleep profile and cortical amyloid load in elderly subjects without dementia. Methods: This cross-sectional study included 143 community-dwelling participants aged ≥70 years (median: 73 years [70-85]; 87 females) with spontaneous memory complaints but dementia-free. Sociodemographic characteristics, health status, neuropsychological tests, sleep, and 18F-florbetapir (amyloid) PET data were collected. The clinical sleep interview evaluated nighttime sleep duration, but also daytime sleep duration, presence of naps, and restless leg syndrome (RLS) at time of study. Validated questionnaires assessed daytime sleepiness, insomnia, and risk of sleep apnea. The cortical standardized uptake value ratio (SUVr) was computed across six cortical regions. The relationship between sleep parameters and SUVr (cut-off ratio>1.17 and tertiles) was analyzed using logistic regression models. Results: Amyloid-PET was positive in 40.6% of participants. Almost 40% were at risk for apnea, 13.5% had RLS, 35.5% insomnia symptoms, 22.1% daytime sleepiness, and 18.8% took sleep drugs. No significant relationship was found between positive amyloid PET and nighttime sleep duration (as a continuous variable, or categorized into <6; 6-7; ≥7 h per night). Logistic regression models did not show any association between SUVr and daytime sleep duration, 24-h sleep duration, naps, RLS, daytime sleepiness, insomnia symptoms, and sleep apnea risk (before and after adjustment for APOEε4 and depressive symptoms). Conclusion: Our study did not confirm the association between amyloid-PET burden, poor sleep quantity/quality in elderly population, suggesting that the interplay between sleep, and amyloid is more complex than described.