Associations of sleep with psychological problems and well‐being in adolescence: causality or common genetic predispositions?

Kristiaan B Van Der Heijden,Marije C.M Vermeulen,Eus J.W Van Someren,Dorret I Boomsma,Charlotte Huppertz,Meike Bartels,Desana Kocevska,Jorien L Treur,Catharina E.M Van Beijsterveldt,Hanna Swaab

doi:10.1111/jcpp.13238

Abstract

BackgroundWhereas short and problematic sleep are associated with psychological problems in adolescence, causality remains to be elucidated. This study therefore utilized the discordant monozygotic cotwin design and cross‐lagged models to investigate how short and problematic sleep affect psychological functioning.MethodsAdolescent twins (N = 12,803, 13–20 years, 42% male) completed questionnaires on sleep and psychological functioning repeatedly over a two‐year interval. Monozygotic twin pairs were classified as concordant or discordant for sleep duration and trouble sleeping. Resulting subgroups were compared regarding internalizing problems, externalizing problems, and subjective well‐being.ResultsCross‐sectional analyses indicated associations of worse psychological functioning with both short sleep and problematic sleep, and cross‐lagged models indicate bidirectional associations. Longitudinal analyses showed that an increase in sleep problems experienced selectively by one individual of an identical twin pair was accompanied by an increase of 52% in internalizing problem scores and 25% in externalizing problem scores. These changes were significantly different from the within‐subject changes in cotwins with unchanged sleep quality (respectively, 3% increase and 5% decrease). Psychological functioning did, however, not worsen with decreasing sleep duration.ConclusionsThe findings suggest that sleep quality, rather than sleep duration, should be the primary target for prevention and intervention, with possible effect on psychological functioning in adolescents.

Highlights

Whereas cross-sectional findings within discordant MZ twin pairs may suggest a causal contribution of short sleep to high INT and EXT, the more powerful longitudinal analyses and the cross-lagged models did not support such interpretation
Consistent with previous literature (Gregory & O’Connor, 2002; Kaneita et al, 2009; Pieters et al, 2015; Sadeh et al, 2014; Shanahan et al, 2014; Wang et al, 2016; Wong et al, 2009), we found that sleep problems were cross-sectionally associated with more INT and EXT and lower subjective well-being (SWB)
Both the cross-sectional discordance analyses and powerful longitudinal analyses within MZ twin pairs are suggestive of a causal contribution of sleep problems to INT and EXT because genetic and shared environmental influences on change over time were ruled out

Summary

Introduction

Sleep problems have consistently been associated with psychopathology in adults (Alvaro, Roberts, & Harris, 2013; Baglioni et al, 2016; Benca, Obermeyer, Thisted, & Gillin, 1992), children (Astill, Van der Heijden, Van IJzendoorn, & Van Someren, 2012; Gregory & Sadeh, 2012; Sivertsen et al, 2015), and adolescents (Brand & Kirov, 2011; Gregory & Sadeh, 2016; Shochat, Cohen-Zion, & Tzischinsky, 2014). After adjustment for genetic and shared environmental influences (since these are identical for identical twins), a shorter sleep duration remained associated with worse self-control and depressive symptoms While these findings are suggestive of a causal effect of sleep duration or quality on adolescent developmental problems, such a conclusion would require support by longitudinal data. One longitudinal study revealed that the association between short sleep duration and mental health status in monozygotic, that is, genetically identical, twin adolescents could not be attributed to shared genetic and environmental factors, providing additional support for a causal contribution (Matamura et al, 2014). Longitudinal analyses showed that an increase in sleep problems experienced selectively by one individual of an identical twin pair was accompanied by an increase of 52% in internalizing problem scores and 25% in externalizing problem scores These changes were significantly different from the within-subject changes in cotwins with unchanged sleep quality (respectively, 3% increase and 5% decrease).

Methods

Results

Conclusion