Subjective-objective sleep discrepancy in older adults : examining the roles of cognition and arousal
Abstract
[EMBARGOED UNTIL 8/1/2024] The difference between subjective (self-report) and objective (e.g., polysomnography, wrist-worn wearables) measures of sleep is referred to as subjectiveobjective sleep discrepancy (SOSD) and is prevalent in insomnia and older adult populations. However, the specific roles of proposed underlying factors that contribute to this discrepancy, such as cognitive functioning and arousal remain unclear due to limited literature examining these factors in older adults with insomnia. Therefore, we examined independent associations between various cognitive domains and arousal (cognitive and physiological) factors on SOSD in older adults and whether insomnia status further moderated these associations. Cognitive and neurologically healthy older adults (N=68, Mage=68.54, 66 percent women) healthy sleepers (N=22) and those with insomnia [N=46; using DSM-5 criteria for insomnia disorder and sleep diary reports of >30 minutes of sleep onset latency (SOL) and/or wake time after sleep onset (WASO) on 3+ nights/7] completed measures of subjective sleep (7-day baseline of sleep diaries), objective sleep [one-night ambulatory polysomnography (PSG)], objective cognition (National Institute of Health Toolbox - Cognition Battery), cognitive arousal (global and pre-sleep cognitive arousal), and physiological arousal [nocturnal cortical arousal, heart rate variability (HRV) at rest and during sleep, and pre-sleep somatic arousal]. For SOSD, the direction (sleep diary variable - PSG variable) and magnitude (|sleep diary variable - PSG variable|) were calculated, using the sleep diary corresponding to the night of PSG recording. Multiple regression analyses examined independent and interactive associations between insomnia status, cognition, arousal, and SOSD for total sleep time (TST), SOL, WASO, and sleep efficiency. Simple slope analyses further evaluated the strength of the associations at different levels of the moderator (healthy sleepers vs. insomnia). Analyses controlled for age, sex, apnea-hypopnea index, and sleep medication usage. Results show that insomnia status moderated associations between specific cognitive domains and arousal factors on both direction and magnitude of SOSD. For cognition, in those with insomnia, worse working memory was associated with worse self-reported TST (i.e., shorter TST) and sleep efficiency (i.e., lower sleep efficiency) than PSG measured. Additionally, in those with insomnia, worse psychomotor processing speed was associated with a smaller magnitude of TST and sleep efficiency SOSD. Further, in those with insomnia, greater pre-sleep cognitive arousal was associated with worse self-reported WASO (i.e., longer WASO) than PSG measured but, interestingly, a smaller magnitude of WASO SOSD. In healthy sleepers, greater physiological arousal at rest (i.e., lower HRV) was associated with a smaller magnitude of SOL discrepancy. Findings suggest that in those with insomnia, poor cognitive functioning related to working memory and psychomotor processing speed as well as greater pre-sleep cognitive arousal may be underlying mechanisms contributing to SOSD. Future studies should investigate this relationship in other objective ratings of sleep (i.e., actigraphy) and other objective measures of arousal (e.g., cortisol), as well as prospectively evaluate whether paradoxical ratings of sleep may be a potential indicator of cognitive decline in insomnia populations.
Degree
M.A.