Factors Associated with Night-to-Night Variability in Sleep Duration During Adolescence

Introduction: Night-to-night variability of sleep variables derived from actigraphy or sleep diary data appears to be a novel risk factor associated with physical and mental health. However, little is known about factors associated with night-to-night sleep variability. The aim of this study was to assess the association of sociodemographic and behavioral factors with night-to-night variability in sleep duration in a general population sample of adolescents.

Methods:A sample of 421 adolescents (age 12-23 y, mean 17.0 +/- 2.3 y; 53.9% male) from the Penn State Child Cohort underwent a 7 consecutive nights actigraphy (GT3X) recording. We calculated within-subject 7-night mean total sleep time and its standard deviation (SD). The SD of total sleep time was used as an index of night-to-night variability in sleep duration. Prior to the actigraphy recording, self-reports on the Children Behavioral Checklist, Morningness-Eveningness Scale, Tanner stage, and sleep patterns were obtained. Linear regression was used to examine the association of potential predictors with night-to-night sleep variability, controlling for age, gender, race, and socioeconomic status (SES).

Results: Older age (β+.110, p=.031), minority status (β=.117, p=.04) and lower SES (β=.142, p=.010) were significantly associated with higher night-to-night sleep variability, whereas gender (β=.075, p=.178) and Tanner stage (β=-.005, p=.927) were not. Independent of these potential confounders, externalizing behaviors (β=.149, p=.008), evening circadian preference (β=-.234, p=.0001), later bedtime on weekdays (β=.134, p=.049), and later waketime on weekends (β=.291, p=.0001) were associated with higher night-to-night sleep variability.

Conclusion: Social, behavioral, circadian, and lifestyle (weekend sleep patterns) factors appear to be associated with night-to-night variability in sleep duration in adolescence. Given the strong association of behavioral and life style patterns with night-to-night sleep duration variability, future studies focus on whether modifying/correcting these aberrant behavioral/life style patterns improve night-to-night variability, a novel risk factor for cardiometabolic morbidity.