Insights from 24-hour actigraphy using functional linear modeling in children with and without ADHD

Sleep problems and alterations in circadian rhythmicity have been widely reported in attention-deficit/hyperactivity disorder (ADHD) patients, but studies covering the full 24h-motor-activity patterns to explore whole-day activity cycles are scarce. Furthermore, correlates and predictors of (hyper-)activity patterns are largely unknown. A subsample of 74 children (6-12 years) with ADHD (n = 35) participating in the multicenter study ESCAlife and n = 39 controls recruited in one center were assessed using 24h-actigraphy for two weeks. Functional linear modeling (FLM) was applied to the actigraphy data and further post-hoc correlational analyses were conducted. FLM revealed no significant differences in 24h-actigraphy profiles between children with and without ADHD, but differentiated between ADHD presentations (combined vs. inattentive) around sleep onset time (around 8:00p.m.) on free days. In the ADHD group, the 24h-actigraphy profiles were associated with measures of the chronotype and early regulation problems in toddler age; in the control group, with daylight exposure. Finding 24h-activity profile differences between ADHD presentations, but not between ADHD and control children as in classical analyses of sleep-onset latency suggests that comprehensive 24h-FLM analyses and focused sleep-related analyses of actigraphy data capture complementary aspects altered in ADHD and its presentations. Another limitation is that the majority of children with ADHD were medicated. However, actigraphy may still prove valuable as a supplementary diagnostic tool in clinical ADHD assessment.