Bedtime in Preschool-Aged Children and Risk for Adolescent Obesity

Abstract

Objective: To determine whether preschool-aged children with earlier bedtimes have a lower risk for adolescent obesity and whether this risk reduction is modified by maternal sensitivity.

Study design: Data from 977 of 1364 participants in the Study of Early Child Care and Youth Development were analyzed. Healthy singleton-births at 10 US sites in 1991 were eligible for enrollment. In 1995-1996, mothers reported their preschool-aged (mean = 4.7 years) child's typical weekday bedtime, and mother-child interaction was observed to assess maternal sensitivity. At a mean age of 15 years, height and weight were measured and adolescent obesity defined as a sex-specific body-mass-index-for-age ≥95th percentile of the US reference.

Results: One-quarter of preschool-aged children had early bedtimes (8:00 p.m. or earlier), one-half had bedtimes after 8:00 p.m. but by 9:00 p.m., and one-quarter had late bedtimes (after 9:00 p.m.). Children's bedtimes were similar regardless of maternal sensitivity (P = .2). The prevalence of adolescent obesity was 10%, 16%, and 23%, respectively, across early to late bedtime groups. The multivariable-adjusted relative risk (95% CI) for adolescent obesity was 0.48 (0.29, 0.82) for preschoolers with early bedtimes compared with preschoolers with late bedtimes. This risk was not modified by maternal sensitivity (P = .99).

Conclusions: Preschool-aged children with early weekday bedtimes were one-half as likely as children with late bedtimes to be obese as adolescents. Bedtimes are a modifiable routine that may help to prevent obesity.

Keywords: epidemiology; longitudinal study; maternal sensitivity; parenting; prospective; sleep.

Figure

Covariate-adjusted predicted prevalence of adolescent obesity by bedtime and maternal sensitivity at preschool-age (∼4.5 years) (SECCYD [n=977]). Children were born in 1991. Adolescent obesity (sex-specific BMI-for-age percentile≥95th) measured at mean=14.9 (SD=0.6) years. Error bars: ±1 SEM. Model adjusted for maternal education, maternal obesity, adolescent age, sex, race/ethnicity and birth weight (8% of cases include imputed data). Bedtime-by-maternal sensitivity interaction not statistically significant (P=0.99).