Both underestimation and overestimation of sleep duration predict mortality in older men with sleep disturbances

While overestimation of sleep duration is associated with increased all-cause mortality in community-dwelling older men, underestimation of sleep duration may have a protective effect. However, this effect could be overwhelmed by concurrent sleep disturbances. Therefore, we examined whether underestimation and overestimation of sleep duration predict mortality risk in older men with sleep disturbances. We analyzed data from 813 participants with sleep disturbance (Pittsburgh Sleep Quality Index > 5) who were not taking medications affecting sleep perception or duration from the Osteoporotic Fractures in Men Sleep Study. Sleep duration discrepancy was quantified using the misperception index (MI). Participants were categorized into high MI (underestimation), intermediate MI (reference), and low MI (overestimation) groups using quartiles. We evaluated the associations of high and low MI with all-cause mortality using Cox proportional hazard model. Overall, 502 deaths (61.7%) were reported during median follow-up of 11.9 years. We found that both high (hazard ratio [HR] = 1.26; 95% confidential interval [CI], 1.01-1.58) and low (HR = 1.42; 95% CI, 1.13-1.78) MI were associated with increased mortality risk after adjusting for demographic, health, and sleep covariates. These findings indicate that both overestimation and underestimation of sleep may be useful for mortality risk assessment in older men with sleep disturbances.