0740 Factor Structure of Sleep in Older Adults Differs when Nearing End-of-Life

Abstract

Abstract Introduction Adults aged 65 and older in the last year of life are at risk for experiencing non-beneficial hospitalizations, adversely effecting their quality-of-life. Reasons for undesired hospitalizations include the need to manage symptoms of progressing chronic illnesses, many of which are exacerbated by poor sleep. Given the complexity of sleep, especially as end-of-life (EOL) nears, it is imperative to understand if the myriad of variables traditionally used to describe sleep can be simplified. Insight gained by such knowledge could be used to guide assessment of sleep at EOL and identification of sleep-related interventions to improve EOL care. Methods Exploratory factor analyses (EFAs) were performed using data from rounds of the National Health and Aging Trends Study (NHATS) that included in-depth sleep modules. One EFA was performed for individuals who died before the next round, and another for those who did not. Ten variables were considered for each EFA, and comprised sleep disorders symptoms, sleep quality, and daytime consequences. Results There were three unique constructs each for the died and not-died groups. Among NHATS participants who had died, sleep experience (>30 mins falling asleep, try falling back asleep, time falling back asleep, sleep hours, and quality of sleep), tiredness (difficulty staying awake, how often nap), and EOL symptoms (sleep meds, breathing, and pain) explained 47.5% of the variability in sleep (n=122). Among NHATS participants who had not died, sleep disorder symptoms (>30 mins falling asleep, try falling back asleep, time falling back asleep, sleep hours, breathing, and quality of sleep), discomfort (pain, difficulty staying awake, how often nap), and sleep medication explained 40.9% of the variability in sleep (n=124). Conclusion Older adults nearing the EOL have different factor structures describing their sleep than those who are not near EOL. Napping and daytime sleepiness as well as pain, medication, and breathing for sleep have important implications for those nearing EOL that could be addressed by palliative care interventions. Support (if any) No support was provided for this study. Dr. Klingman is funded, in part, by NINR Award Number R01 NR018979. Dr. Sullivan is funded, in part, NIA Award Number R03 AG067159.