Vo, Tien2021-09-242021-09-242021-07https://hdl.handle.net/11299/224550University of Minnesota Ph.D. dissertation. July 2021. Major: Social and Administrative Pharmacy. Advisor: Angeline Carlson. 1 computer file (PDF); x, 196 pages.Sleep-disordered breathing (SDB) or sleep apnea, a common disorder, is characterized by repeated pauses or reductions in breathing during sleep. A number of studies, primarily in younger or middle-aged populations, have evaluated the association of SDB and healthcare utilization. However, previous studies were limited by use of cross-sectional or case-control study designs and inadequate control of potential confounders. Sleep efficiency, defined as the percentage of time in bed spent sleeping, is a key measure of sleep health and has been shown to decrease with advancing age along with total sleep time, another important sleep parameter, defined as total hours per night spent sleeping while in bed.While some studies have examined sleep efficiency as a predictor of adverse health outcomes and conditions, there is a paucity of research that has considered sleep efficiency as an outcome measure. To address current gaps in research presented above, the goals of the proposed study are to achieve the following expected outcomes: First, we will estimate the prevalence of sleep-disordered breathing and determine the association of SDB with subsequent measures of health care utilization and costs in U.S. community-dwelling older men. The findings will provide a clearer understanding of the impact of sleep-disordered breathing on healthcare costs and inpatient and post-acute care utilization, and possibly warrant future intervention studies that would have public health impact to determine whether treatment of sleep-disordered breathing lowers these measures of healthcare burden. Second, using standard logistic regression, we will examine and identify factors that are associated with incident reduced sleep efficiency in U.S. older community-dwelling men and women. The findings will provide insights on potential modifiable predictors of incident reduced sleep efficiency and guide design of future intervention studies. Third, we will use machine-learning methods through random forests to identify factors of importance in explaining incident reduced sleep efficiency in U.S. older community-dwelling men and women. Ultimately, this research proposal will improve our understanding of the determinants of the development of incident reduced sleep efficiency in older men and women, and quantify the impact of sleep-disordered breathing on total healthcare costs and utilization in older men.enActigraphy and PolysomnographyHealthcare Costs and UtilizationMachine LearningRandom ForestsSleep EfficiencySleep-disordered BreathingThesis or Dissertation