Abstract:Abstract Introduction Racial/ethnic residential segregation (RRS) likely contributes to sleep health disparities. For instance, nighttime lighting for surveillance and policing efforts that vary by neighborhood racial composition likely contributes to differences in exposure to artificial light at night (ALAN), which disrupts melatonin production and sleep. Yet, the RRS-ALAN-sleep pathway is understudied. Methods To investigate RRS and ALAN as mediators of racial/ethnic differences in sleep among US adults, we geographically linked National Health Interview Survey (2011-2017) data to 2012 American Community Survey 5-year population estimates and outdoor ALAN data from the 2010-2011 US Defense Meteorological Satellite Program’s Operational Linescan System. The local Getis-Ord Gi* was used to categorize RRS (low [z-score< 0], medium [0<=z-score<=1.96], high [z-score>1.96]). ALAN was divided into quintiles. Short sleep duration (< 7-hours) and poor sleep quality were self-reported. We used adjusted, survey-weighted robust Poisson regression to estimate prevalence ratios (PRs) and 95% confidence intervals (CIs); tested for effect modification by race/ethnicity; and assessed mediation using the Sobel’s test. Results Among 126,539 adults (mean±SE age=46±0.1 years), RRS prevalence was 35% low, 43% medium, and 22% high. RRS and poor sleep were higher among minoritized adults (non-Hispanic Black (NHB), Mexican Latino, non-Mexican Latino, and Asian); ALAN was higher among non-Hispanic White (NHW) adults. High RRS was associated with a lower prevalence of ALAN (Q5 vs. Q1: PR=0.59 [95% CI:0.44-0.80]), short sleep (PR=0.94 [0.91-0.97]), and poor sleep quality (e.g., PR-trouble falling asleep =0.94 [0.90-0.98]), with no differences by race/ethnicity. ALAN was not associated with short sleep (sleep quality data unavailable). RRS explained differences in short sleep between both NHB and Mexican vs. NHW adults (e.g., NHB: PR-total effect=1.30 [1.26-1.33]; PR-direct effect=1.39 [1.34-1.45]; PR-indirect effect=0.93 [0.89-0.98]) before but not after adjustment for ALAN (PR-indirect effect=0.97 [0.87-1.08]). Conclusion RRS was associated with lower ALAN and non-poor sleep. RRS attenuated sleep duration disparities among NHB and Mexican adults only prior to adjustment for ALAN. RRS and ALAN may be contrasting modifiable contributors to sleep disparities among US adults. Support (if any) The findings/conclusions in this research are those of the authors and do not necessarily represent the views of the Research Data Center, National Center for Health Statistics, or CDC.

0251 Racial Residential Segregation and Outdoor Artificial Light at Night: Potential Contributors to Sleep Disparities