Abstract
Introduction
Alzheimer’s disease and related dementias (ADRD) are growing public health concerns, and poor sleep may represent a modifiable risk factor. However, there is limited research on insomnia as a predictor of subsequent performance in different cognitive domains and mechanisms that might underlie domain-specific associations. The current study examined: (1) which insomnia symptoms predicted performance across five cognitive domains 14 years later, and (2) whether depressive symptoms and/or vascular diseases mediated these associations.
Methods
Participants included 2,496 adults aged 51+ in the Health and Retirement Study. Insomnia symptoms in 2002 (i.e., “baseline”) were quantified by four self-reported items on frequency of trouble falling asleep, nighttime awakenings, early awakenings, and feeling rested upon awakening. Cognition was assessed in 2016 as part of the Harmonized Cognitive Assessment Protocol and operationalized with five factor scores corresponding to episodic memory, executive function, language, visuoconstruction, and processing speed. Multiple regressions examined associations between baseline insomnia symptoms and subsequent cognitive performance, controlling for sociodemographics and baseline global cognitive performance. Mediation models tested whether associations were explained by self-reported depressive symptoms and/or vascular diseases (i.e., hypertension, heart disease, diabetes, and/or stroke) in 2014, controlling for baseline values.
Results
Only trouble falling asleep in 2002 was associated with cognition in 2016. Specifically, more frequent trouble falling asleep predicted poorer episodic memory, executive function, language and processing speed performance, but not visuoconstruction. These associations were mediated by depressive symptoms and vascular diseases in 2014 for all domains except episodic memory; only depressive symptoms mediated the association involving memory. After accounting for these mediators, direct effects of trouble falling asleep remained for episodic memory, executive function and language, but not processing speed.
Conclusion
Difficulty with sleep initiation may be more consequential for later-life cognition than other insomnia symptoms. Depressive symptoms and vascular diseases may partially drive these associations. We speculate that sleep-onset insomnia could mean less total sleep, immune dysfunction, or endocrine effects that worsen mood, vascular health, and cognition. Remaining associations indicate that additional research is needed to characterize other mechanisms through which sleep initiation problems could contribute to later impairments in frontal and temporal cognitive systems, which are implicated early in ADRD.
Support (if any):
P1-434: Is age-related variance in episodic memory mediated by processing speed or executive function?
