O070 Longitudinal associations of sleep microarchitecture with future cognitive function in middle-aged and older men from a community-based cohort study

STUDY OBJECTIVES: Prospective studies examining associations of sleep microarchitecture with future cognitive function predominantly recruited from small samples with relatively short follow-up. Therefore, this study examined sleep microarchitecture predictors of cognitive function (visual attention, processing speed, and executive function) after 8–10 years in community-dwelling middle-aged and older men. METHODS: Of 433 Florey Adelaide Male Ageing Study participants who underwent home-based polysomnography (2010–2011), 157 completed baseline and follow-up cognitive testing. Whole-night F4-M1 sleep EEG recordings were processed following artefact exclusion, and quantitative EEG features obtained using validated algorithms. Trail-making tests A (TMT-A) and B (TMT-B) were administered at baseline (2007–2010) and follow-up (2018–2019) examinations. Linear regression models were adjusted for baseline obstructive sleep apnea, demographic, biomedical, and behavioural factors, and cognitive performance. RESULTS: At baseline, participants were aged mean (SD) 58.9 (8.9) years with normal cognition. In unadjusted models only, lower REM sleep delta and higher alpha power were associated with worse TMT-A performance at follow-up (delta, B= -0.02, 95% CI [-0.04, -0.001], p=0.042; alpha, B=0.08, 95% CI [0.01, 0.15], p=0.024). Lower overall and slow sleep spindle density in N2 and N3 were associated with worse TMT-A performance at follow-up (all p<0.05). The association of higher fast spindle density in N3 with worse TMT-B performance (B =1.06, 95% CI [0.13, 2.00], p=0.026) was not significant after adjustment for baseline cognition. CONCLUSIONS: In this sample of community-dwelling middle-aged and older men with normal baseline cognition, sleep microarchitecture parameters were not independently associated with cognitive function after 8–10 years.