Sleep microstructure and cognitive trajectories in healthy aging: Insights from a longitudinal cohort study

Background Age-related changes in sleep microstructure may contribute to cognitive decline, but their value as markers of cognitive trajectories in healthy aging is unclear. We examined associations between baseline sleep microstructure and both concurrent cognition and ten-year cognitive change in older men. Methods Men from the Metropolit 1953 Danish Male Birth Cohort completed cognitive assessment at two time points a decade apart (2010–2013 and 2020–2022; n = 206 at baseline, n = 106 at follow-up). Polysomnography was performed at baseline only (n = 176); microstructure analyses (spindles, slow waves, EEG coherence) were restricted to participants with adequate NREM sleep (n = 82). Cross-sectional and longitudinal associations were assessed using Spearman and partial Spearman correlations with Benjamini–Hochberg correction. Results Over ten years, processing speed (TMT-B: +6.1 s) and symbol conversion (SDMT: −2.3) declined significantly. Higher sleep efficiency was associated with better cognitive maintenance, whereas increased wake after sleep onset and an increased arousal index correlated with poorer memory and learning. A higher proportion of N1 sleep was linked to poorer executive function, whereas N2 and N3 were positively associated with cognitive outcomes. Fast spindle amplitude and frequency, slow-wave characteristics, and N2 and REM coherence correlated with processing speed, attention, and global cognition. Conclusions Baseline sleep microstructure, including spindle, slow-wave dynamics and EEG coherence, was associated with cognitive aging and may provide candidate markers of cognitive resilience or decline. These exploratory findings support further longitudinal studies to determine their predictive value for neurodegenerative disease.