Abstract
BACKGROUND: Previous studies have amply demonstrated shorter self-reported sleep duration for Black patients compared to White patients. Evidence for disparities in objectively measured sleep is still mixed. Our objective is to assess disparities in objectively measured sleep parameters between races/ethnicities using data from an academic medical center in the US. MATERIALS AND METHODS: We used data extracted through text mining from sleep reports of in-lab polysomnography (PSG) studies conducted for adults aged 18 years or older at the University of Virginia Health System between 2010 and 2021. All studies with total sleep time (TST) <1 h were excluded. In multivariable analyses, we controlled for age, sex, body mass index, marital status, season, health insurance, comorbidities, and use of medications that may interfere with sleep in 1 year prior to the sleep study. RESULTS: The study sample included 5,331 patients of whom 69% were non-Hispanic (NH) White, 23% NH Black, 4% Hispanic/Latinx, and 3% other or unknown. They were 57 ± 16, 52 ± 14, 49 ± 14, and 54 ± 14 years old and 45%, 30%, 33%, and 41% male, respectively. Average TST was 342 ± 84 min, sleep efficiency 74%, apnea-hypopnea index (AHI) 15 ± 17, with 69% having obstructive sleep apnea (36% mild; 22% moderate; 12% severe). After adjusting for covariates, Black patients had shorter N3 sleep by 10 min (95% CI = -11.9, -7.6) and longer REM sleep by 7 min (95% CI = 4.8, 8.6) compared to NH White patients. We did not find significant differences in TST, AHI, and sleep efficiency between racial/ethnic groups. CONCLUSIONS: Our objective data does not show consistently unfavorable sleep measures for racial and ethnic minorities. More research is needed to better understand how much of these differences are due to underlying physiology vs. social/environmental factors.