Abstract
OBJECTIVES: To test the hypothesis that sleep disturbances are independently associated with frailty status in older men. DESIGN: Cross-sectional analysis of prospective cohort study. SETTING: Six U.S. centers. PARTICIPANTS: Three thousand one hundred thirty-three men aged 67 and older. MEASUREMENTS: Self-reported sleep parameters (questionnaire); objective parameters of sleep-wake patterns (actigraphy data collected for an average of 5.2 nights); and objective parameters of sleep-disordered breathing, nocturnal hypoxemia, and periodic leg movements with arousals (PLMAs) (in-home overnight polysomnography). Frailty status was classified as robust, intermediate stage, or frail using criteria similar to those used in the Cardiovascular Health Study frailty index. RESULTS: The prevalence of sleep disturbances, including poor sleep quality, excessive daytime sleepiness, short sleep duration, lower sleep efficiency, prolonged sleep latency, sleep fragmentation (greater nighttime wakefulness and frequent, long wake episodes), sleep-disordered breathing, nocturnal hypoxemia, and frequent PLMAs, was lowest in robust men, intermediate in men in the intermediate-stage group, and highest in frail men (P-for-trend < or =.002 for all sleep parameters). After adjusting for multiple potential confounders, self-reported poor sleep quality (Pittsburgh Sleep Quality Index > 5, multivariable odds ratio (MOR)=1.28, 95% confidence interval (CI)=1.09-1.50), sleep efficiency less than 70% (MOR=1.37, 95% CI=1.12-1.67), sleep latency of 60 minutes or longer (MOR=1.42, 95% CI=1.10-1.82), and sleep-disordered breathing (respiratory disturbance index > or =15, MOR=1.38, 95% CI=1.15-1.65) were each independently associated with higher odds of greater frailty status. CONCLUSION: Sleep disturbances, including poor self-reported sleep quality, lower sleep efficiency, prolonged sleep latency, and sleep-disordered breathing, are independently associated with greater evidence of frailty.