545 Actigraphic sleep parameters and anxiety symptoms among community-dwelling older adults

Abstract

Abstract Introduction A number of cross-sectional studies have found that elevated levels of anxiety are associated with poor sleep among healthy older adults, but most have used self-reported sleep measures. We investigated the longitudinal association between objectively measured sleep (by wrist actigraphy) and subsequent change in anxiety symptoms in this population. Methods We studied 555 community-dwelling older adults (mean age 72.52±7.35, 77.48% white, 53.15% women) in the National Social Health and Aging Project (NSHAP) study who completed 3 nights of wrist actigraphy at wave 2 (2010–2011) and the Hospital Anxiety and Depression Scale at waves 2 and 3 (2015–2016). Actigraphic sleep parameters were averaged across nights and included: total sleep time (TST; minutes), percent sleep (%), wake after sleep onset (WASO; minutes), and sleep fragmentation. Change in anxiety was calculated as the difference between anxiety scores at wave 3 and wave 2. Results After adjusting for age, race, sex, education, body mass index, number of medical conditions, depression symptoms, and anxiety scores at wave 2, we found no significant associations between any actigraphic sleep parameter and subsequent change in anxiety symptoms (all p ≥ 0.390). Additional analyses revealed no significant cross-sectional associations at wave 2 (p ≥ 0.390). Conclusion We found no evidence for an association between actigraphic sleep and anxiety symptoms, or change in anxiety symptoms, in community-dwelling older adults. Additional studies using clinical anxiety disorder diagnoses are needed to evaluate the extent to which objectively measured sleep disturbance predicts clinically significant anxiety in older adults. Support (if any) American Academy of Sleep Medicine Foundation (#223-BS-19).