1068 DIFFERENCES IN SELF-REPORTED AND OBJECTIVE LONG SLEEP TIME: THE MULTI-ETHNIC STUDY OF ATHEROSCLEROSIS

As with short sleep duration, long sleep has been associated with adverse outcomes, however, it is not known whether long sleep duration is intrinsically detrimental or a manifestation of a co-morbid conditions such as sleep disordered breathing, poor sleep and chronic illness. Moreover, previous data has largely relied on a self-reported sleep duration. In this study, we compare the differences among self-reported long sleepers (≥ 9 hours), self-reported non-long sleepers (<9 hours) and objective long sleepers ≥ 9 hours, measured by actigraphy. We analyzed data from the Multi-Ethnic Study of Atherosclerosis in participants with completed sleep questionnaires, polysomnography and actigraphy (N=2127). We compared age, socio-demographics, body mass index (BMI), presence of diabetes, Epworth Sleepiness Scale score, apnea hypopnea index (AHI), sleep latency, wake after sleep onset (WASO), total and awake time in bed and sleep efficiency among subgroups of subjective and objective long sleep compared to those without long sleep in bivariate analyses. The prevalence of self-reported long sleep duration was 23% (498 of 2127), however only 7% (35 of 498) of these actually had ≥9 hours of sleep on actigraphy (2% of sample). Self-reported long sleepers were older (72 ± 9 vs 69 ± 9 years p<0.01), more often female (60% vs 52% p<0.01) with a higher prevalence of diabetes (26% vs 17% p<0.01) but did not differ by BMI, AHI, sleepiness, compared to those reporting <9 hours of sleep. By actigraphy, self-reported long sleepers only slept 7.0 ± 1.6 hours, had greater WASO (42.0 ± 16 vs. 35.8 ± 15 minutes) and greater total time in bed (7.8 ± 1.66 vs. 7.1 ± 1.40 hours) and awake time in bed (49 ± 20 vs. 43 ± 17 minutes) but no difference in sleep efficiency or sleep latency minutes than those reporting <9 hours of sleep. Self-report of long sleep was much more common than objectively measured long sleep. Only 7% of self-reported long sleepers have objectively assessed long sleep. Self-reported long sleepers had more time awake in bed and greater WASO. These results suggest that inferences regarding the health effects of long sleep duration cannot be reliably made by questionnaire alone. MESA Sleep NHLBI RO1L098433, NIH HL098433