# Racial-Ethnic Residential Segregation and Sleep Health among US Adults: Associations by Race and Ethnicity, Sex/Gender, and Neighborhood-Level Poverty

## TL;DR

This study finds that racial-ethnic residential segregation is linked to better sleep health for some groups, but effects vary by race, gender, and neighborhood poverty.

## Contribution

The study provides new insights into how segregation affects sleep health and identifies key modifiers like sex and poverty.

## Key findings

- High RRS was associated with lower prevalence of short and long sleep durations and higher restorative sleep.
- Men experienced stronger associations with less trouble falling asleep compared to women.
- Race-ethnicity-by-sex/gender and neighborhood poverty modified sleep outcomes without consistent patterns.

## Abstract

Although racial-ethnic residential segregation (RRS) is hypothesized to contribute to sleep disparities by concentrating poverty and impairing sleep among minoritized racial-ethnic groups, feelings of belonging within relatively homogenous neighborhoods may be protective against poor sleep. Yet, empirical studies are sparse.

To investigate RRS–sleep health associations and determine potential modifiers among US adults, we linked National Health Interview Survey data (2011–2017) to 2012 and 2017 American Community Survey census tract-level data. We used the local Getis-Ord Gi* statistic to categorize RRS (high, medium, low [reference]). Using survey-weighted, Poisson regression with robust variance, we estimated prevalence ratios (PRs) and 95% confidence intervals (CIs) for self-reported sleep health measures. We also performed Wald tests for interactions by race-ethnicity, sex/gender, race-ethnicity-by-sex/gender intersectional category, and neighborhood-level poverty.

Among 126,539 participants (mean age ± SE = 46 ± 0.1 years), high RRS was most common among non-Hispanic (NH)-Black (38%), followed by NH-Asian and non-Mexican Latine (34%), Mexican Latine (30%), and NH-White adults (17%). Across races-ethnicities and sexes/genders (both p-interaction > 0.05), high vs. low RRS was associated with a 6% lower prevalence of short sleep duration (< 7-h: PR = 0.94 [95% CI:0.91–0.97]), an 11% lower prevalence of long sleep duration (> 9-h: PR = 0.89 [0.80–0.99]), and a 2% higher prevalence of restorative sleep (PR = 1.02 [1.01–1.04]). Associations with a lower prevalence of trouble falling asleep were stronger among men vs. women. Race-ethnicity-by-sex/gender group membership and neighborhood-level poverty modified associations with sleep duration and quality without consistent patterns.

RRS was associated with more favorable sleep health among US adults with variation by key modifiers (e.g., sex). Strategies that leverage potentially protective social factors while promoting equitable resources across diverse neighborhoods may help address sleep health disparities.