Bedtime, body mass index and obesity risk in preschool-aged children

Abstract

Background: Although sleep duration is a risk factor for obesity in young children, less is known about other aspects of sleep health, including bedtime, on obesity risk.

Objective: To determine whether bedtime is associated with body mass index (BMI) z-score or obesity risk in children ages 2 to 5 years, and to determine if associations are independent of sleep duration.

Methods: Cohort analyses were undertaken using three early life obesity prevention trials (POI, INSIGHT, Healthy Beginnings) and a longitudinal cohort study (HOME). Bedtime was assessed by questionnaire and BMI through clinical measurement between 2 and 5 years in 1642 children. Adjusted regression models examined whether BMI z-score and obesity (BMI z-score ≥ 2) were associated with bedtime, nocturnal sleep time and 24-hour sleep time. A discrete mixture model categorized children into bedtime trajectory groups across time points.

Results: Bedtime was inconsistently associated with BMI z-score. Although each hour later of bedtime was associated with greater odds of obesity at ages 3 (OR; 95% CI: 1.05; 1.003, 1.10) and 5 (1.35; 1.08, 1.69) years, odds were attenuated after adjustment for nocturnal or 24-hour sleep time. Longer nocturnal sleep duration at 2 years was associated with lower odds of obesity (OR 0.90; 0.86, 0.94), as was longer 24-hour sleep duration at 3 years in girls (0.70; 0.62, 0.78). BMI z-score and odds of obesity were not significantly different between 'early to bed' and 'late to bed' trajectory groups.

Conclusions: Timing of bedtime appears inconsistently related to obesity in young children, possibly via influencing overall sleep duration.

Keywords: body mass index; child; obesity; preschool; sleep.

Conflict of interest statement

CONFLICT OF INTEREST

The authors declare no conflicts of interest.

© 2020 World Obesity Federation.

Figures

FIGURE 1

Median bedtimes by age (bars are 25th and 75th percentiles) for each trajectory group found using discrete mixture modelling (n = 1112). ‘Early to bed’ group had n (%) = 871 (78.3), ‘Late to bed’ group had n (%) = 241 (21.7)