Sleep architecture and dementia risk in adults: an analysis of 5 cohorts from the Sleep and Dementia Consortium

STUDY OBJECTIVES: Poor sleep may play a role in the risk of dementia. However, few studies have investigated the association between polysomnography (PSG)-derived sleep architecture and dementia incidence. We examined the relationship between sleep architecture and dementia incidence across five US-based cohort studies from the Sleep and Dementia Consortium. METHODS: Percent of time spent in stages of sleep (N1, N2, N3, rapid eye movement sleep), wake after sleep onset, sleep maintenance efficiency, apnea-hypopnea index, and relative delta power were derived from a single night home-based PSG. Dementia was ascertained in each cohort using its cohort-specific criteria. Each cohort performed Cox proportional hazard regressions for each sleep exposure and incident dementia, adjusting for age, sex, body mass index, antidepressant use, sedative use, and APOE e4 status. Results were then pooled in a random effects model. RESULTS: The pooled sample comprised 4657 participants (30% women) aged ≥ 60 years (mean age was 74 years at sleep assessment). There were 998 (21.4%) dementia cases (median follow-up time of 5 to 19 years). Pooled effects of the five cohorts showed no association between sleep architecture and incident dementia. When pooled analysis was restricted to the three cohorts which had dementia case ascertainment based on DSM-IV/V criteria (n = 2374), higher N3% was marginally associated with an increased risk of dementia (hazard ratio (HR): 1.06; 95%CI: 1.00-1.12, per percent increase N3, p = .050). CONCLUSIONS: There were no consistent associations between sleep architecture measured and the risk of incident dementia. Implementing more nuanced sleep metrics and examination of associations with dementia subtypes remains an important next step for uncovering more about sleep-dementia associations.