Insomnia, early and late rising are associated with small hippocampal volume and large white matter hyperintensity burden.

Abstract

Background: Sleep disturbances have been associated with an increased risk of dementia. The mechanisms remain unclear, although neurodegenerative and vascular pathways are potentially involved. Hence, our study aims to investigate the relationships between several clinical sleep and polysomnographic features and volumes of hippocampus (indicative of neurodegeneration) and white matter hyperintensities (WMH) (reflecting vascular processes).

Methods: In this cross-sectional study, 678 participants aged 65-80 from the French population-based ESPRIT cohort with MRI-measured hippocampus and/or WMH volumes were included. Self-reported sleep data were collected at baseline, and 176 participants underwent ambulatory polysomnography (PSG). We performed multivariable logistic regression to assess associations between sleep characteristics and hippocampal and WMH volumes.

Results: Participants' median age was 70.7 years (Q1-Q3 = 67.8-74.0), with 52.4% being women. Early (≤ 6 am; odds ratio (OR) = 2.03, 95% confidence interval (CI) = 1.17;3.53) and late (> 8 am; OR = 2.14, 95%CI = 1.33;3.43) rising times were associated with low hippocampal volume. Early rising time (OR = 2.06, 95%CI = 1.24;3.43) and insomnia symptoms (OR = 1.84, 95%CI = 1.18;2.86 for 1 symptom, OR = 1.91, 95%CI = 1.18;3.09 for 2-3 symptoms) were associated with large WMH volume, whereas late bedtime (≥ 11 pm; OR = 0.56, 95%CI = 0.39;0.80) was associated with low WMH volume. Based on PSG data, higher rapid-eye movement (REM) sleep percentage (OR = 0.70, 95%CI = 0.50;0.96) was associated with low WMH volume, with similar trends for long sleep bouts duration, N3 and REM sleep durations (p = 0.05 to 0.07). Conversely, higher N2 sleep percentage (OR = 1.69, 95%CI = 1.09;2.62), longer NREM sleep bouts (OR = 1.46, 95%CI = 1.02;2.09), and higher periodic leg movements index (OR = 1.55, 95%CI = 1.02;2.26) were associated with large WMH volume. However, no PSG parameter associations remained after false discovery rate correction.

Conclusions: Distinct associations between sleep characteristics and hippocampal and WMH volumes were observed, highlighting the important relationships between sleep, sleep timing and brain structure.