Bedtime and evening light exposure influence circadian timing in preschool-age children: A field study

Abstract

Light exposure and sleep timing are two factors that influence inter-individual variability in the timing of the human circadian clock. The aim of this study was to quantify the degree to which evening light exposure predicts variance in circadian timing over and above bedtime alone in preschool children. Participants were 21 children ages 4.5-5.0 years (4.7 ± 0.2 years; 9 females). Children followed their typical sleep schedules for 4 days during which time they wore a wrist actigraph to assess sleep timing and a pendant light meter to measure minute-by-minute illuminance levels in lux. On the 5th day, children participated in an in-home dim-light melatonin onset (DLMO) assessment. Light exposure in the 2 h before bedtime was averaged and aggregated across the 4 nights preceding the DLMO assessment. Mean DLMO and bedtime were 19:22 ± 01:04 and 20:07 ± 00:46, respectively. Average evening light exposure was 710.1 ± 1418.2 lux. Children with later bedtimes (lights-off time) had more delayed melatonin onset times (r=0.61, p=0.002). Evening light exposure was not independently associated with DLMO (r=0.32, p=0.08); however, a partial correlation between evening light exposure and DLMO when controlling for bedtime yielded a positive correlation (r=0.46, p=0.02). Bedtime explained 37.3% of the variance in the timing of DLMO, and evening light exposure accounted for an additional 13.3% of the variance. These findings represent an important step in understanding factors that influence circadian phase in preschool-age children and have implications for understanding a modifiable pathway that may underlie late sleep timing and the development of evening settling problems in early childhood.

Keywords: Bedtime; Children; Circadian phase; DLMO; Light; Melatonin; Preschool.

Fig. 1

Scatterplots of the association between bedtime and DLMO (A) and of evening light exposure and DLMO after controlling for bedtime (B; partial correlation, residuals displayed). Lines show the slope of the regression line.

Fig. 2

Visual depiction of variance in circadian timing predicted by bedtime and evening light exposure (n=21; ages 4.5–5.0 years). Forty-nine percent of the variance in dim light melatonin onset (DLMO) timing was unaccounted for but may be due to additional factors such as genetics, light history, wake time and morning light.