Does context matter? A multilevel analysis of neighborhood disadvantage and children's sleep health

Abstract

Objectives: To determine how demographic, socioeconomic, and neighborhood characteristics are associated with bedtimes among US kindergarteners.

Design: Parents reported bedtimes of their children as well as personal, household, and residential characteristics via interviews in the Early Childhood Longitudinal Study-Kindergarten (ECLS-K) Class of 1998-1999. The ECLS-K links individual households to US Census tracts.

Setting: A random selection of 1,280 schools and surrounding communities in the US.

Participants: A random selection of 16,936 kindergarteners and their parents.

Measurements: The 2 outcomes were regular and latest weekday bedtimes of kindergarteners. Through a series of nested multilevel regression models, these outcomes were regressed on individual- and neighborhood-level variables, including race/ethnicity, sex, family type, household income, mother's educational attainment, neighborhood disorder, and several additional neighborhood characteristics.

Results: Models showed significant (P < .05) bedtime disparities by race/ethnicity, sex, family income, and mother's educational attainment. Additionally, models tended to indicate that kindergarteners from disadvantaged neighborhoods experienced later bedtimes than children from more advantaged areas. Neighborhood characteristics accounted for a portion of racial/ethnic differences, suggesting that bedtime disparities are partly rooted in disparate environmental conditions.

Conclusions: Reducing disparities in childhood sleep may require programs that target not only children and their parents, but also the communities in which they reside.

Keywords: Bedtime; Census tracts; Children; ECLS-K; Multilevel models; Neighborhoods; Sleep; United states.