Actigraphy- and Polysomnography-Measured Sleep Disturbances, Inflammation, and Mortality Among Older Men

ABSTRACT Objectives To evaluate whether objectively measured sleep characteristics are associated with mortality risk independent of inflammatory burden and comorbidity. Methods The Osteoporotic Fractures in Men Sleep Study (conducted in 2003–2005) included community-dwelling older men ( n = 2531; average [standard deviation {SD}] age = 76.3 (5.5) years). Sleep measures from in-home polysomnography and wrist actigraphy and assessments of serum inflammatory markers levels (C-reactive protein, interleukin-6, tumor necrosis factor α, tumor necrosis factor α soluble receptor II, and interferon-γ) were obtained. Vital status was ascertained over an average (SD) follow-up of 7.4 (1.9 SD) years. Results Three of the seven main sleep measures examined were independently associated with greater inflammatory burden. Mortality risk associated with prolonged (≥10% total sleep time) blood oxygen desaturation and short (<5 hours) sleep duration was attenuated to nonsignificance after adjusting for inflammatory burden or medical burden/lifestyle factors. Severe blood oxygen desaturation (adjusted hazard ratio [aHR] = 1.57, 95% confidence interval [CI] = 1.11–2.22), sleep fragmentation (aHR = 1.32, 95% CI = 1.12–1.57), and a lower percentage of sleep in rapid eye movement (aHR per SD = 0.90, 95% CI = 0.93–0.97) were independently associated with mortality. Conclusions Short sleep duration and prolonged blood oxygen desaturation were independently associated with inflammatory burden, which attenuated associations between these sleep characteristics and mortality. Medical and life-style factors also substantially attenuated most sleep-mortality associations, suggesting complex relations between sleep, inflammation, and disease. Sleep fragmentation, severe blood oxygen desaturation, and the percentage of sleep time in rapid eye movement were independently related to mortality risk. Future studies with repeated measures of mediators/confounds will be necessary to achieve a mechanistic understanding of sleep-related mortality risk.