Aim: Using national Australian time-diary data, we aimed to empirically determine sleep duration thresholds beyond which children have poorer health, learning, quality of life, and weight status and parents have poorer mental health.
Design/setting: Cross-sectional data from the first three waves of the Longitudinal Study of Australian Children.
Participants: A nationally representative sample of 4983 4-5-year-olds, recruited in 2004 from the Australian Medicare database and followed biennially; 3631 had analyzable sleep information and a concurrent measure of health and well-being for at least one wave.
Exposure: At each wave, a parent completed 24-h time-use diaries for one randomly selected weekday and one weekend day, including a "sleeping/napping" category.
Outcomes: Parent-reported child mental health, health-related quality of life, and maternal/paternal mental health; teacher-reported child language, literacy, mathematical thinking, and approach to learning; and assessed child body mass index and girth.
Results: Linear regression analyses revealed weak, inconsistent relationships between sleep duration and outcomes at every wave. For example, children with versus without psychosocial health-related quality of life problems slept slightly less at 6-7 years (adjusted mean difference 0.12 h; 95% confidence interval 0.01-0.22, p = 0.03), but not at 4-5 (0.00; -0.10 to 0.11, p = 1.0) or 8-9 years (0.09; -0.02 to 0.22, p = 0.1). Empirical exploration using fractional polynomials demonstrated no clear thresholds for sleep duration and any adverse outcome at any wave.
Conclusions: Present guidelines in terms of children's short sleep duration appear misguided. Other parameters such as sleep timing may be more meaningful for understanding optimal child sleep.
Keywords: Child; Development; Epidemiological studies; Reference values; Sleep; Time diary.
Copyright © 2015 Elsevier B.V. All rights reserved.