Abstract 4140013: Sleep Disparities Across Demographics and Cardiometabolic Disorders in the NIH All of Us Fitbit Database

Background: Prior research has noted disparities in sleep duration among demographic groups and those with cardiometabolic disorders. However, these are mostly based on self-reported data. The NIH All of Us Fitbit database offers a new method for objective and reliable sleep assessment. Goals: The study aimed to objectively assess sleep duration using the All of Us Fitbit database across various demographic variables and cardiometabolic disorders. Methods: All of Us participants with at least one year of Fitbit data were identified. Fitbit’s "minutes asleep" parameter was extracted daily over the first year of Fitbit use and averaged. The average total minutes asleep (TMA) was compared across self-reported age, sex and race groups. For those individuals who also shared their electronic health record (EHR) data, TMA was compared between those with and without hypertension, diabetes, and sleep apnea. T-test and ANOVA were used for comparisons. Results: The first year of Fitbit data for 13,039 participants (51 [16]* years, 69% female, 82% White) was analyzed, with sleep information available for 330 [104] days (90% complete data). TMA decreased with age, with the 18-44, 45-64, and 65+ groups averaging 366 [64], 348 [72], and 339 [85] minutes respectively (p<0.0001). Men and women had a small difference in TMA (342 [74] vs 351 [77] mins respectively, p<0.0001). Individuals who self-reported as Black or African-American had significantly lower TMA (304 [68] mins) compared to other groups (White: 353 [76] mins, Asian: 334 [67] mins, Other: 344 [69] mins; p<0.0001). Among 8587 individuals who also shared their EHR data, lower TMA was observed in those with, compared to those without, prevalent hypertension (332 [83] vs 353 [76] mins), diabetes (320 [82] vs 349 [78] mins), and sleep apnea (328 [77] vs 350 [79] mins) respectively (all p<0.0001). *mean [SD] Conclusion: Fitbit data analysis identified variations in sleep duration across age, race, and presence of cardiometabolic disorders. Objective sleep assessment using wearables could provide more reliable insights into sleep disparities and their associations with cardiometabolic outcomes.