0176 fMRI Neural Activation During Working Memory After 5-Nights of Sleep Restriction in Adolescents with ADHD Traits

Abstract Introduction Persistent sleep loss is common among adolescents. Its impact on cognitive functioning is well documented, particularly in working memory. We hypothesize that children with attention-deficit/hyperactivity-disorder (ADHD) may be particularly vulnerable to such effects. Here we examine the impact of 5-nights of at-home sleep restriction on the neural correlates of working memory in young adolescents with and without ADHD traits. Methods Twenty-five peripubertal adolescents (12.36±0.88 years, 14F) were characterized by Conners-3-Parent ADHD-Index [>=/< 50% tile] as either high (ADHDy; n=11) or low (ADHDn; n=14). All participants completed an fMRI-monitored working memory task after two counterbalanced conditions: 5-nights of sleep optimization (SO; 10h TIB set to habitual risetime) and 5-nights of sleep restriction (SR; 7.5h TIB, equally delaying bedtime and advancing risetime). At least two nights of stabilization preceded both conditions. fMRI sessions consisted of two 7-minute runs of alternating 2-back and 0-back working memory blocks. Voxel-wise 2x2 linear mixed effects models (3dLME) investigated condition [SR vs. SO], group [ADHDy vs. ADHDn], and interaction effects on working memory activation (2-back>0-back). Significance was set to p<.005, k=30 voxels. Results Our actigraphy monitored protocol significantly reduced sleep period time by 20% (SR= 7.39±0.43h vs. SO= 9.19±0.39h; [t(24)=17.66, p<.001; d=3.53]) and total sleep time by 17% (SR= 6.79±0.51h vs. SO= 8.20±0.60h; [t(24)=12.9, p<.001; d=2.58]). At SO, fMRI 2-back blocks were associated with increased activation in task-on areas, such as the bilateral dorsal lateral prefrontal, anterior cingulate, and superior parietal cortices, and decreased activation in default-mode task-off areas, including the precuneus and medial prefrontal cortex. Of these regions, significant condition-by-group effects indicated that sleep loss increased activation in the bilateral superior parietal cortex (MNIx,y,z: [40, -48, 68], k=172) and bilateral cuneus/precuneus (MNIx,y,z: [16, -84, 48], k=1446), only in the ADHDy group. Conclusion These data are the first to indicate that ADHD status may influence how sleep loss affects working memory in the adolescent brain. Sleep loss increased activation of both task-on (e.g., superior parietal) and task-off (e.g., precuneus) regions in the ADHDy group. We speculate this pattern may reflect less efficient neural processing after sleep loss. Ongoing data collection will probe sources of inter-individual variability. Support (if any) R01HD103655; P20GM139743