The Relationship Between ADHD Symptom Severity and Sleep Disturbance Using Accelerometer and the Subjective Sleep...

There is limited evidence regarding the relationship between the severity of ADHD symptoms and the level of sleep disruption in early adolescent children. This study aimed to better understand this relationship by studying early adolescents from the Adolescent Brain Cognitive Development (ABCD) Study at the 2-year follow-up visit (mean age 12.0, n = 4,414).

The "Attention Problems" scale of the Child Behavior Checklist (CBCL) was used to measure the severity of ADHD symptoms for study participants. Sleep was assessed both subjectively with the Sleep Disturbance Scale for Children (SDSC) via parent report and objectively by wrist-worn accelerometer assessment of sleep duration.

Total sleep disturbance showed a moderate positive correlation with ADHD symptoms (r = .41, p < .001). In contrast, there was only a very small negative correlation between accelerometer-measured sleep duration and ADHD symptoms (r = -.05, p < .001). Multivariate linear regression with total SDSC score as the dependent variable, adjusting for demographics, medication use, and puberty stage, showed a positive association of ADHD symptoms with total SDSC score (β = .41, 95% CI [0.38-0.44]; p < .001). A longitudinal linear mixed-effects model analysis with total SDSC score as the dependent variable also showed a positive association with ADHD symptoms with total SDSC score (β = .34, 95% CI [0.32-0.35]; p < .001). In contrast, multivariate linear regression with total sleep duration (in minutes) as the dependent variable showed only a very small negative association (β = -.03, 95% CI [-0.0-0.00]; p < .05).

Our study shows that increased severity of ADHD symptoms is associated with worse sleep disturbance as measured by subjective parent reports; however, there is a very weak relationship between ADHD symptoms and objectively measured total sleep duration. The contrasting results are consistent with prior literature, and could be due to unmeasured variables such as deep sleep percentage or non-REM EEG delta wave power that are not captured by actigraphy. Alternatively, there is the possibility of mono-informant bias, as both the CBCL and SDSC are parent-reported, and parents may perceive relationship between worse ADHD symptoms and worse sleep disturbances, though the difference is objectively negligible.