Paudel, Misti Lyn2014-12-232014-12-232014-09https://hdl.handle.net/11299/168278University of Minneswota Ph.D. dissertation. September 2014. Major: Epidemiology. Advisors: Adviser: Kristine E. Ensrud, M.D., MPH Co-Adviser: Russell Luepker, M.D., MS. 1 computer file (PDF); ix, 141 pages, appendices A-N.Complaints of insufficient or poor sleep are highly prevalent in older adult populations, with up to 50% reporting symptoms of fragmented sleep, difficulty falling asleep, early awakening and short sleep duration. In this dissertation, I focus on improving our understanding of the epidemiology of sleep disturbances in older adults by addressing three important gaps in the literature.Aim 1: I evaluated the ability of a clinical tool, the STOP-BANG screening questionnaire, to detect obstructive sleep apnea(OSA) in 2,953 older men. The STOP-BANG identified 88.4% of men as having a high likelihood of OSA, resulting in a large number of false positives (sensitivity=94%, specificity=12.7, PPV=18.6%). Results suggest that the STOP-BANG has limited clinical value in a community-dwelling older male population. Aim 2: Little is known about the intra-individual variability of sleep, especially in older adult populations who have a greater burden of sleep complaints. In a cohort of 2,804 older men, significant intra-individual variability in sleep measures was observed, and the strongest independent associative factors with greater sleep variability included race, living alone, smoking, antidepressant use, benzodiazepine use, depression, greater BMI and greater comorbidity burden. Aim 3: Sleep disturbances are associated with comorbid medical conditions, however the extent to which sleep disturbances independently impact inpatient healthcare utilization is not well understood. Data from the Study of Osteoporotic Fractures (SOF), linked to Medicare claims and Kaiser Permanente encounters among 2,103 older women was used to address this question. During 3 years of follow-up, 55% of the cohort was hospitalized at least once. Significant associations between sleep disturbances and greater odds of being hospitalized (31-72% increased odds) were largely explained by greater comorbidity burden, depression and health-related factors in multivariable adjusted models. Findings suggest that sleep disturbances are not independently associated with greater inpatient health-care utilization in older women.enAgingHealth care utilizationMedicareSleepEpidemiologyThesis or Dissertation