01014 Sleep and Patient-Reported Symptoms in Older Adults during Critical Illness Recovery: Comparisons by Sleep Medication Exposure and Chronotype

Abstract Introduction Older adults recovering from critical illness experience fragmented nighttime sleep, daytime napping, and symptom burden that may influence the early recovery period through transition from intensive care unit (ICU) to acute care. It is unclear whether sleep medication administration or individual chronotype are associated with better sleep. Our primary aims were to compare sleep and symptom profiles: 1) across sleep medication exposure groups, and 2) between chronotypes. Our exploratory aim was to identify factors associated with perceived sleep during the acute care observation period. Methods Our analysis included 27 older adult ICU survivors after transfer to acute care who consented to ≥24 hours of actigraphy/sleep observation. The primary exposure was sleep medication: no sleep medications, melatonin-only, or other (e.g., trazodone, quetiapine, etc). Chronotype was dichotomized as morning or intermediate/afternoon via the Composite Scale of Morningness. Sleep and symptom profiles included: actigraphy-measured/objective nighttime sleep (10:00PM-6:00AM) and daytime naps (6:00AM-10:00PM), daily perceived sleep ratings (0-10), and patient reported symptoms (PROMIS T-scores). Analyses used one-way ANOVA for medication groups, independent samples t-tests for chronotype groups, and multiple linear regression for average perceived sleep quality. Results Across medication groups, there were no significant differences in nighttime sleep, daytime naps, perceived sleep ratings, or patient-reported symptoms. However, intermediate/afternoon chronotypes reported worse sleep related impairment (65.23 versus 53.14; p<.01) and sleep disturbance (61.60 versus 52.41; p<.01). Among those still hospitalized on study observation day 5, intermediate/afternoon chronotypes tended to log more objective nighttime sleep and daytime naps than morning chronotypes, although this comparison was non-significant (p=.09). In multiple regression (R²=0.31), sleep disturbance was a negative correlate of average perceived sleep ratings (β=-0.58; p=.004); chronotype, depression, and objective sleep were not associated. Conclusion In this cohort of hospitalized older adult ICU survivors after post-ICU transition to acute care, sleep medication exposure was not associated with improved sleep or symptom profiles. Day-by-day perceived sleep varied among those who received sleep medications and did not align with objective sleep. Findings suggest that future work should evaluate chronotherapeutic, non-pharmacologic interventions to improve sleep. Support (if any) NIH/NIA K23AG078448; American Association of Critical-Care Nurses; University of Washington de Tornyay Center for Healthy Aging