Associations between self-reported sleep patterns and health, cognition and amyloid measures: results from the Wisconsin Registry for Alzheimer’s Prevention

• DOI: 10.1093/braincomms/fcad039
• Publication Date: 2023-02-24T03:59:42Z
• Authors: Lianlian Du; Rebecca Langhough; Bruce P Hermann; Erin Jonaitis; Tobey J Betthauser; Karly Alex Cody; Kimberly Mueller; Megan Zuelsdorff; Nathaniel Chin; Gilda E Ennis; Barbara B Bendlin; Carey E Gleason; Bradley T Christian; David T Plante; Rick Chappell; Sterling C Johnson
• Keywords: Epworth Sleepiness Scale; Cognitive Decline

Abstract Previous studies suggest associations between self-reported sleep problems and poorer health, cognition, Alzheimer’s disease pathology dementia-related outcomes. It is important to develop a deeper understanding of the relationship these complications disturbance, modifiable risk factor, in late midlife, time when may be accruing. The objectives this study included application unsupervised machine learning procedures identify distinct subgroups persons with problematic association concurrent measures mental physical cognition PET-identified amyloid. Dementia-free participants from Wisconsin Registry for Prevention (n = 619) completed questionnaires including Insomnia Severity Index, Epworth Sleepiness Scale Medical Outcomes Study Sleep Scale. K-means clustering analysis identified discrete problem groups who were then compared across health outcomes (e.g. depression, self-rated insulin resistance), cognitive composite indices episodic memory executive function and, subset, Pittsburgh Compound B PET imaging assess amyloid burden. Significant omnibus tests (P < 0.05) followed pairwise comparisons. Mean (SD) sample baseline assessment age was 62.6 (6.7). Cluster three groups: healthy sleepers [n 262 (42.3%)], intermediate 229 (37.0%)] poor 128 (20.7%)]. All comparing demographics significant except age, sex apolipoprotein E e4 carriers; group worse than one or both other on all measures, complaints. had higher average body mass index, waist–hip ratio homeostatic model resistance. After adjusting covariates, also performed composites working memory. There no differences PET-based Sensitivity analyses indicated that while different approaches resulted assignments some (predominantly group), between-group patterns consistent. In conclusion, characteristics sizable minority (20.7%) exhibiting characteristics, exhibited poorest indicating substantial multi-morbidity; not associated estimates. Precision-based management related factors provide an opportunity early intervention could serve delay prevent clinical impairment.