Longitudinal Associations of Sleep Quantity and Depressive Symptoms From Adolescence Through Young Adulthood in a National Sample

Background: Sleep plays an important role in cognitive and emotional function and is an essential determinant of health and well-being. Sleeping less is associated with higher levels of depressive symptoms (DS) during adolescence, but the causal direction of this association is not clear. It is also not clear whether the association of DS and sleep is consistent in adolescence and young adulthood. This study uses national longitudinal data to assess prospective effects of sleep and DS from adolescence through young adulthood.

Methods: Data were from the National Longitudinal Study of Adolescent Health, a four-wave nationally representative longitudinal study from adolescence through young adulthood. Structural equation modeling was used to estimate a cross-lag model testing the directionality of the association of sleep quantity (hours of sleep per night) and DS (measured using a brief version of the Centers for Epidemiological Studies-Depression–CES-D—scale) across four waves.

Four models were estimated and compared: 1) continuity model, in which continuity in sleep and DS were modeled with no cross-lag associations between them, 2) sleep-driven model, testing cross-lag associations from sleep to subsequent DS, 3) depression-driven model, testing cross-lag associations from DS to subsequent sleep quantity, and 4) full cross-lag model, allowing cross-lag associations from sleep to DS and from DS to sleep.

Results: The depression-driven model showed the best fit with the data (X²(13)=80.57, CFI=.97, TLI=.92, RMSEA=.03). Wave 1 DS were negatively associated with wave 2 sleep quantity (B=-.20, p=.002); Wave 2 DS were positively associated with Wave 3 sleep quantity (B=.24, p=.001). There were no significant prospective effects of sleep quantity on DS.

Conclusions: This study is the first to assess directionality in the association of DS and sleep quantity across adolescence to young adulthood in a national sample. Results indicate that DS drive sleep quantity, rather than the reverse. There was some evidence for developmental specificity of the direction of the relationship: in adolescents, the effect of DS on sleep was negative; those with higher rates of DS at wave 1 reported less sleep at wave 2. From adolescence into young adulthood, the association was positive; those with higher rates of DS at wave 2 reported more sleep at wave 3. Additional analyses will test subgroup differences in these associations and mediators of the association of DS and sleep in order to probe mechanisms underlying developmental differences in this association and inform potentially developmentally specific interventions addressing DS and sleep disturbance.