Youth’s bedtime regularity mediates the association of depression and anxiety with negative attention bias

Introduction: Attention biases towards negative information and sleep problems are core clinical features in anxiety and depressive disorders. We studied if sleep patterns, which were readily malleable, were associated with such biases among depressed and anxious individuals. Methods: A youth sample recruited from local universities (n=188, age ranged from 17–24) were administered the Structured Clinical Interview for DSM-IV disorders, (47% have a lifetime history of depressive and/or anxiety disorder). All participants wore an actiwatch and completed a sleep diary for 5 days and then completed an affective go/no-go task (AGNG) and other neurocognitive assessment on the 6th day. Their sleep difficulty (e.g. sleep latency), and regularity of sleep timing (e.g. standard deviation of 5-day bedtime) were coded as indicator of sleep-wake behaviors with the discrimination index (d’) in the AGNG as indicator of attention bias. Results: Depressed and anxious individuals had a lower d’ in the AGNG, F(1,176)=6.280, p=.013, 2=.038 than healthy participants. There was a significant correlation between d’ on the AGNG with regularity of bedtime, r(168)=.225, p=.024. Structural equation model with depression and anxiety as predictor, bedtime regularity as mediator and d’ on AGNG as dependent variable achieved a very good fit, CFI=1.000, RMSEA<0.001, SRNR<.001. There was no significant direct effect (B=.032, Standard Error=.211, p=.880) but a significant indirect effect of depression and anxiety status in predicting d’ on AGNG through its effect on bedtime regularity. B=.180, Standard Error=.127, p=.031. Conclusion: Our findings support the mediating role of sleep of the associations between depressive and anxiety disorder with attention bias. Given the high prevalence of depressive and anxiety disorders and sleep problems among youth and the prevalence of attention bias in psychopathology, further longitudinal studies are warranted in investigating the causal interrelationships among these variables. Support (If Any): Funder: General Research Fund (#18619616), Research Grant Council, HKSAR