0367 The Independent and Interactive Effects of Insomnia Symptoms and Short Sleep on Sleep Physiology

Abstract Introduction The presence of subjective insomnia complaints with objective short sleep duration (total sleep time [TST] < 6 hours/night) has been associated with more severe health outcomes compared to insomnia or shortened sleep alone. The differential and possibly interacting effects of insomnia and shortened sleep on sleep physiology have not yet been comprehensively characterized. Methods 1,014 participants from the Sleep Apnea Global Interdisciplinary Consortium (SAGIC) without sleep apnea (apnea-hypopnea index < 5 events/hour) were included. Participants were categorized as having insomnia or good sleepers based on self-reported symptoms and as short (< 6 hours/night) or normal (≥6 hours/night) sleepers based on in-laboratory TST. Analysis of variance models adjusted for age and sex assessed the effects of insomnia, shortened sleep, and their interaction on traditional sleep architecture (percent of TST in each sleep stage), EEG power metrics, and novel odds ratio product (ORP; indexes sleep depth ranging from 0 [deep sleep] to 2.5 [full wakefulness]) metrics (average ORP in each sleep stage and ORP-9 [speed with which sleep deepens following arousal]). Results Neither insomnia status, sleep duration, nor their interaction had significant effects on the percent of TST spent in any sleep stage. The interaction between insomnia status and sleep duration was not significant for any ORP or EEG metrics. However, individuals with insomnia demonstrated lighter sleep (higher NREM ORP [p=0.02, η2=0.01], REM ORP [p =0.002, η2=0.02], and ORP-9 [p=0.0002, η2=0.03]), cortical hyperarousal (higher beta power in 14-20 Hz [p=0.0003, η2=0.03] and 20-35 Hz [p=0.001, η2=0.02]), and higher sigma power (12-16 Hz [p=0.04, η2=0.01]) compared to good sleepers. Short sleepers also demonstrated lighter REM sleep (higher ORP [p=0.0004, η2=0.03]) and cortical hyperarousal (higher beta power in 14-20 Hz [p< 0.0001, η2=0.04] and 20-35 Hz [p< 0.0001, η2=0.04]) as well as lower delta power (1-4 Hz [p=0.0009, η2=0.02]) and theta power (4-7 Hz [p< 0.0001, η2=0.05]) relative to normal sleepers. Conclusion While insomnia and shortened sleep did not influence traditional sleep architecture, they were differentially associated with alterations in sleep physiology. These more granular metrics may be useful in future investigations of potentially separable mechanisms that underlie adverse health consequences in insomnia versus shortened sleep. Support (if any)