ABSTRACT
OBJECTIVE: Self-reported "sleep quality" often is assessed in epidemiologic studies. However, the bases for variation in sleep quality is not fully understood. We quantified the extent to which subjective sleep quality was related to sleep disorders and sleep characteristics among 795 African American adults. METHOD: Between 2012 and 2016, participants underwent home sleep apnea testing and 1-week actigraphy (estimating sleep duration, efficiency, fragmentation, latency). Sleep quality, insomnia and restless legs syndrome symptoms, sleepiness, and physician diagnosis of sleep disorders were self-reported. We fit linear regression models to determine the extent to which subjective and objective sleep measures as well as depressive symptoms and anxiety were related to subjective sleep quality. RESULTS: After adjustment for covariates, worse sleep quality scores were associated with insomnia and restless legs syndrome symptoms, sleep apnea, physician diagnosis of a sleep disorder, and actigraphy-based fragmented sleep, lower sleep efficiency, and shorter sleep duration. Insomnia symptoms explained the most variance in subjective sleep quality, 21%. Other sleep measures each explained 3% to 7% and psychosocial factors explained 8% to 9% of the variance in subjective sleep quality after adjustment for confounders. CONCLUSIONS: The weak associations of sleep quality with sleep disorders and objectively measured sleep disturbances are consistent with concepts of "sleep health" as a multidimensional construct. Sleep quality is a patient-centered outcome that provides unique information over objective measurements of sleep disturbances.