Abstract Introduction Cross-sectional and prospective data in adolescents indicate that sleep and circadian rhythm characteristics are related to substance use and related problems. Experimental rodent and human data suggest sleep deprivation and circadian misalignment alter reward sensitivity and impair impulse control. Adolescents subjected to early school start times suffer sleep loss and circadian misalignment, and thus may be at particular risk for substance use through reward- and/or impulsivity-related mechanisms. In a sample of adolescents in middle or high school, we tested whether a chronotherapeutic intervention altered reward sensitivity and/or impulsivity relative to a control condition. Methods Sixty-eight 13-15 year-old middle/high school students (34 female) completed a 2-week baseline protocol before completing a randomly-assigned 2-week chronotherapeutic intervention (n=33) or a sleep-monitoring control condition (n=35). Participants assigned to the intervention maintained a stable wake-up time, 30 minutes morning bright light (via Re-Timer glasses), 2 hours evening dim light (via blue-blocking glasses), and a 1.5-hour earlier weeknight bedtime. Sleep was assessed via actigraphy during 2-week baseline and experimental periods. Each concluded with an overnight sleep lab visit during which circadian phase was assessed with the dim light melatonin onset (DLMO), while reward and impulsivity were assessed via self-report (BIS/BAS; UPPS-P) and behavioral tasks (Balloon Analogue Risk Task [BART]; Cued Go/No-Go Task [CGNG]). Mixed models accounted for biological sex. Results The intervention advanced DLMO by 34 minutes (B=-0.57, p=0.019) and extended weeknight total sleep time by 35 minutes (B=0.58,p=0.005) relative to the control. Behavioral impulsivity (CGNG) worsened in the control condition relative to the intervention (B=0.08,p=0.023). Across both groups, advances in DLMO correlated with increases in behavioral reward motivation (BART; B=-3.18, p=0.014). Conclusion Our preliminary findings suggest that a novel chronotherapeutic intervention may alter reward sensitivity and impulsivity in complex ways. However, most of the sample had not initiated substance use, and it remains unclear whether effects would generalize to adolescents with greater substance use, or if a longer-term intervention and/or follow-up period might reveal further effects. Future studies should consider longer-term interventions in substance-using adolescents to test impact on reward and impulsivity processes as well as substance use itself. Support (if any) P50DA046346 (MPIs: McClung, Buysse; Project Director: Hasler)

Abstract Introduction Identifying biological risk-markers for bipolar disorder (BD) remains challenging. Meta-analyses suggest individuals at-risk for BD exhibit significant variability in daily routines and irregular sleep patterns, potentially indicating circadian rhythm dysregulation. We investigated whether biological circadian measures predicted symptoms of mania over time in young people at clinical high-risk for BD. Methods In an ongoing study, 92 participants aged 16-24 (M= 21.9, SD=2.11) were recruited across a spectrum of lifetime subthreshold mania symptoms (Mood Spectrum Measure-Lifetime; MOODS-SR-L). At baseline, salivary dim-light melatonin onset (DLMO) assessed circadian phase on weekday (Thursday) and weekend (Sunday) nights; measures of circadian phase (average of weekday and weekend DLMO), phase instability (weekend-weekday DLMO difference) were derived. The post-illumination pupil response (PIPR) measured melanopsin-driven light responsivity by comparing blue to red 200 ms light stimuli at 6sec (PIPR6) and 10–40sec (PIPR30) after stimulus onset. Mania (Altman Self-Rating Mania Scale; ASRM) and depressive (Patient Health Questionnaire-9; PHQ-9) symptoms were evaluated monthly over up to 3-years follow-up. As high and/or variable levels of manic symptoms over time are associated with BD risk, mixed effect models assessed average levels of mood symptoms over follow-up and generalized linear mixed effect models with a dispersion parameter examined mood symptom variation over follow-up. All models varied for age, sex, psychiatric medication use, family bipolar history, retinal irradiance, and photoperiod. Results Greater circadian phase instability was associated with higher average mania symptoms over follow-up (βinstability=0.50, p=0.022). Both greater circadian phase instability (βinstability=0.05, p< 0.001) and higher melanopsin responsivity (βPIPR6=2.38, p< 0.001) were associated with more variability in mania symptoms over time. Additionally, later circadian phase (βphase=0.05, p< 0.001) and phase instability (βinstability=0.10, p< 0.001) were associated with more variability in depression symptoms over time. Conclusion Our interim results indicate that elevated melanopsin responsivity may be a promising biological marker of mania risk, whereas later circadian phase specifically relates to worsening depression over time. Circadian phase instability predicted longitudinal mood lability more broadly. These data provide insight into potential mechanisms underpinning the mood stabilizing effects of social rhythms therapy and dark therapy for BD; future experimental or intervention studies should evaluate causal relationships. Support (if any)

Abstract Introduction Light Therapy (LT) is a promising non-pharmacological treatment for depression, however, the mechanisms supporting its therapeutic benefits remain unclear. Preclinical models indicate that light modulates mood through melanopsin-containing of retinal ganglion cells (mRGCs). mRGCs are maximally sensitive to blue light and minimally sensitive to red light, and directly convey light signals from the retina to brain structures involved in threat and reward processing. Using within-scanner light exposures, we examined the degree to which melanopsin-engaging blue (vs. red light and darkness) light modulated regional metabolism within brain regions supporting threat and reward processing in adults with depressive symptoms. Methods A total of 33 young adults (18-30yr, 24.94±3.17yr; 20 Female) with elevated depressive symptoms (Patient Health Questionnaire-9>5) completed 1 week of a stable sleep schedule followed by an MRI assessment and pupillometry assessment of melanopsin-driven light responsivity (post illumination pupil response). During the MRI protocol, participants underwent pseudo-continuous arterial spin labeling to assess cerebral blood flow (CBF) during dark, blue, and red light exposures lasting approx. 5 minutes; the order of red and blue light was counterbalanced. A mixed effects model evaluated CBF differences in threat (amygdala, insula, ventromedial prefrontal cortex [vmPFC]) and reward (ventral striatum[VS], medial prefrontal cortex[mPFC]) network regions of interest, adjusting for age, sex, and depression severity. Results Light condition impacted CBF in the VS (F=5.32, p=0.008), mPFC (F=4.41, p=0.017), vmPFC (F=4.63, p=0.014), and insula (F=6.02, p=0.004). Activation was greater in red light versus dark in the VS (p=0.014), mPFC (p=0.020), vmPFC (p=0.017), and insula (p=0.004). There were no significant differences between dark and blue light, or red and blue light, contrary to our predictions. Melanopsin responsivity moderated effects of blue vs red light on CBF (F=5.29, p=0.03), such that greater melanopsin responsivity was associated with greater mPFC CBF under blue but not red light conditions (b=38.7, p=0.05). Conclusion Among individuals with elevated depressive symptoms, bright light exposure may modulate metabolism threat and reward-related brain regions implication in depression pathophysiology in complex ways. Sensitivity of the melanopsin system could play a role in the degree to which blue light modulates affective brain function. Support (if any)

Abstract Introduction Adolescence and young adulthood are key periods for the emergence of bipolar disorder (BD). Instability in sleep and RAR increases BD risk, with recent evidence showing that its impact varies depending on pre-existing mania vulnerability. Yet it remains unclear whether this vulnerability extends to the brain. This question is critical because if pre-existing mania vulnerability heightens neural sensitivity to sleep and RAR instability, it could exacerbate disruptions in emotion-regulation circuits, a core process implicated in BD, and thereby increase susceptibility to its onset. White matter organization in fronto-limbic pathways, which supports emotion regulation and is sensitive to sleep/RAR disruption, offers a promising neural substrate for probing this vulnerability. Methods Participants (16–24y; N=112), recruited across a continuum of mania vulnerability (MOODS-SRL - Mania), completed 14 days of actigraphy and a neuroimaging session. Diffusion MRI was used to derive Neurite Orientation Dispersion and Density Imaging, focusing on the Orientation Dispersion Index (ODI) of emotion-regulation tracts (cingulum bundle, forceps minor, uncinate fasciculus). Clinician-rated mania and depression were assessed at baseline and six-months. Variability in sleep duration and sleep onset was calculated, along with the Circadian Function Index (CFI), a composite measure of circadian activity rhythm robustness. We tested whether associations between actigraphy-derived sleep/RAR metrics and ODI were moderated by mania vulnerability at baseline, and whether ODI metrics at baseline predicted 6-month mood symptoms. Results Lower CFI (greater circadian instability) was associated with higher ODI (greater white matter disorganization) in the uncinate fasciculus (β=-0.22, P=0.018) among individuals with higher mania vulnerability. Higher uncinate fasciculus ODI predicted greater mania symptoms at follow-up and fully mediated the CFI-mania association (β=-0.10,95%CI[-0.14,-0.02]). No effects emerged for depressive symptoms. Sleep duration and onset variability were not associated with ODI (P>0.05). Conclusion Our findings suggest that circadian instability specifically impacts the uncinate fasciculus in individuals at elevated risk for BD, and this microstructural vulnerability predicts future increases in mania but not depressive symptoms. These results highlight a potential pathway linking rhythm disruption to emerging mania and underscore circadian rhythm stabilization as a promising target for early intervention aimed at protecting white matter integrity and reducing BD risk. Support (if any) National Institute of Mental Health (R01MH124828; PI: Soehner).

Objective:Mania symptoms in youth predict poor long-term mental health outcomes, yet their developmental trajectories and associated risk factors remain unclear. Methods:Leveraging data from the Adolescent Brain Cognitive Development Study (N=10,474; 9-10 years at baseline; 48% female, 65% white), we used latent growth mixture models to identify trajectories of mania symptoms across two years in early adolescence. We used multinomial logistic regressions to examine associations between trajectories and risk factors across mental and physical health, cognition, and family/environmental domains. Results:We identified three trajectories: Low (58%), Moderate (32%), and High/Variable Mania Symptoms (10%). All mental health symptoms (depression, attention deficit hyperactivity disorder, conduct and oppositional defiant disorders and anxiety), two physical health factors (sleep disturbances, irritable bowel syndrome symptoms), one cognitive factor (verbal learning impairment), and three family/environmental factors (trauma, parent- and youth-reported family conflict) significantly differentiated between all three trajectories, reflecting incremental increases in risk factor severity with increasing mania symptoms. Other physical, cognitive, family and environmental factors were also associated with more severe mania symptom trajectories. Conclusions:More severe mania symptom trajectories in early adolescence are associated with multiple mental, physical, cognitive, family and environmental risk factors, underscoring the need for comprehensive risk prediction approaches in youth.

Abstract Introduction Cognitive control is a core regulatory process implicated in externalizing behaviors (e.g., anger, impulsivity), and it is typically measured by average accuracy. Less work has examined trial-to-trial variability, which may capture instability in control processes. Sleep is another key regulator of affect and behavior, yet its role in shaping links between cognitive control and externalizing behavior is not well understood. We tested whether cognitive control variability (CCV) and accuracy predicted two facets of externalizing behavior, anger and impulsivity, and whether sleep duration and efficiency moderated these associations. Methods Participants (N = 120; baseline, N = 89; 6-month; Mage = 21.7) were recruited across a spectrum of mental health symptoms. They completed two weeks of sleep monitoring via daily diary and actigraphy. Cognitive control was assessed with the Multi-Source Interference Task, from which two indices were derived: trial-to-trial variability in reaction time (CCV) and mean accuracy. Anger (past 7 days) and impulsivity (negative urgency) were assessed with self-report scales. Analyses tested whether sleep duration and efficiency moderated associations between CCV/accuracy and anger/impulsivity, covarying age, sex, mental health symptoms, and mean reaction time. Results For anger, higher CCV predicted greater symptoms, particularly when diary-based sleep duration was shorter (β=-0.08, p=.02). By 6 months, moderation effects diminished, but CCV continued to show direct associations with anger. For impulsivity, accuracy, not CCV, was the primary predictor. At baseline, actigraphy-based sleep efficiency moderated this link, such that higher efficiency increased the protective effect of accuracy. (β=--1.15, p=.01) By 6 months, higher accuracy predicted lower impulsivity across sleep indices. Sleep duration and efficiency showed no independent main effects. Conclusion Different aspects of cognitive control mapped onto distinct facets of externalizing. Variability in control signaled risk for anger, especially under shorter sleep, consistent with anger as a proximal, state-like outcome sensitive to short-term sleep and moment-to-moment instability. Accuracy was most relevant for impulsivity, aligning with impulsivity as a more trait-like disposition that is less tied to immediate sleep but is predicted by accuracy over time. Variability and accuracy thus index distinct pathways to externalizing, with sleep shaping how those pathways are expressed. Support (if any) R01MH124828

Adolescence is a period of distinct maturational changes in sleep physiology. Age-related trends in sleep physiology have been captured using laboratory-based polysomnography, a method limited by logistical burden and high cost. We tested the ability of the Dreem3 sleep EEG headband to replicate established age effects in sleep physiology from late childhood through early adulthood. Typically developing youth (N = 100, 9–26 years) completed 3–4 consecutive nights of at-home sleep recording. We estimated age-related trends across eight macro-architecture and 15 micro-architecture variables with known age effects, and conducted exploratory analyses of 24 additional variables. Dreem3 replicated established age trends, including increases in non-rapid eye movement (NREM) stage 2%, and decreases in N3%, time in bed, NREM delta and theta power with increasing age. Exploratory analysis revealed age effects in twelve additional variables, including decreases in spindle activity with increasing age. Sleep EEG wearables offer an accessible way to characterize sleep physiology development.

Study objectives:Adolescence is a period of distinct maturational changes in sleep characteristics. Historically, age trends in sleep physiology have been captured using laboratory-based polysomnography (PSG). However, multiple challenges associated with PSG, including logistical issues, budgetary constraints and ecological validity questions, limit large-scale use. The current study aims to address these challenges by using the Dreem3 headband to measure sleep at home and replicate well-established age-related trends in sleep physiology from late childhood through early adulthood. Methods:100 typically developing youth (9-26 years) wore a sleep electroencephalography (EEG) device (Dreem3) for 3-4 consecutive nights at home. Sleep EEG data were processed using the Luna pipeline. We used linear mixed models to estimate age-related trends across 8 macro-architecture and 15 micro-architecture variables previously found to be associated with age, and explored age relationships in 24 additional macro- and micro-architecture variables. Results:At-home sleep studies using Dreem3 replicated established age trends in sleep macro- and micro-architecture, including decreases in percent time spent in non-rapid eye movement (NREM) stage 3 (N3%) sleep and decreases in NREM delta power with increasing age. Exploratory analysis revealed age effects in seven other variables, including decreases in integrated slow spindle activity and NREM cycle duration with increasing age. Conclusion:Sleep EEG wearables may offer an accessible way to characterize sleep physiology development in large cohorts, setting the stage for understanding how deviations from normative age patterns may put young people at risk for adverse outcomes.

Mania is associated with circadian and reward dysregulation, and there are strong reciprocal relationships between the circadian and reward systems. To better detect and prevent mania, it is essential to characterize potentially abnormal biobehavioral relationships between the circadian and reward systems that may confer risk for mania. Light signals conveyed through melanopsin-containing retinal ganglion cells are one of the strongest influences on the biological clock. We examined the extent to which melanopsin-driven light responsivity (via pupillometry) is associated with reward dysregulation in young people at-risk for mania. Ninety-five participants aged 16-24yr (M=21.82, SD=2.02) spanning a spectrum of mania vulnerability (MOODS-SR-Lifetime, MOODS) completed a 24-hr lab visit. Testing included melanopsin-driven pupil responsivity (post-illumination pupil response, PIPR), reward-based aggression (Point Subtraction Aggression Paradigm, PSAP), and reward motivation (Effort Expenditure for Rewards Task, EEfRT). PIPR was estimated at 10-40sec (PIPR30) after the light stimulus, calculated as a percent of baseline. Poisson regression models assessed PIPR’s associations with reward-related outcomes and moderation by lifetime mania risk (MOODS mania score), adjusting for age, sex assigned at birth, past-week depression and mania symptoms, psychotropic medication use, time spent awake, photoperiod, and MOODS depression score. Johnson-Neyman intervals were used to examine the range of significant moderation (interaction) effects. There was a significant interaction (RR=1.45. p< 0.001) between melanopsin-driven light sensitivity (PIPR30) and mania risk (MOODS-Mania) for reward-related aggression (PSAP percent steals). A Johnson-Neyman test showed that lower PIPR30 was associated with greater PSAP percent steals at low/moderate mania vulnerability (MOODS-Mania values < 16) but this relationship was reversed for individuals with higher levels of mania vulnerability (MOODS-Mania> 23). There was not a significant association between PIPR30 and reward motivation (EEfRT % hard choices; RR=0.14, p=0.643) nor a significant moderating effect of mania risk (RR=2.00. p=0.238). Our interim findings indicate that, as mania risk increases, greater melanopsin-driven pupil responsivity is associated with greater reward-related aggression. This finding may reflect abnormal associations between the circadian and reward systems among individuals with elevated vulnerability to mania that, when exacerbated, contribute to irritability and aggressive behavior associated with mania/hypomania.

It is well known that children with attention deficit hyperactivity disorder (ADHD) have an elevated risk for poor sleep health. Sensory over-responsivity is also common for children with ADHD potentially impacting the ability for children to transition to sleep. Current interventions do not incorporate sensory-specific theories to decrease hyperarousal level at bedtime to support sleep health. This pilot study trials a novel sensory-based intervention targeting bedtime cognitive and physiological hyperarousal. Children (6-13 yrs) with ADHD and their caregivers are being recruited to participant in a 3-week pilot intervention trial (total goal n=30). Biophysiological measures of arousal (e.g. electrodermal activity, physical activity) and sleep (actigraphy) are captured using the Empatica EmbracePlus and daily sleep and emotion diaries are completed. Caregivers complete a manualized gentle pressure massage and mindfulness protocol called the “Power Down”. Intervention feasibility and acceptability are measured through qualitative interviews and questionnaires. Preliminary efficacy is measured through change in sleep and bedtime arousal measures at the end of the intervention trial (2 weeks long) compared to a 1-week baseline. Recruitment is on-going. Currently, one participant has completed the three-week study (8 years old, male) and one is half-way through the study (11 years old, male). Both participants and their caregivers identified transitioning to sleep as the primary difficulty related to sleep health. Both participant’s caregivers reported that the Power Down intervention was acceptable, appropriate, and feasible after initial training (all agree to completely agree, 5-pt Likert scale). Data collection from the wearable device was feasible, with participants wearing the device for 22/25 days and 13/14 days of data collection. The Power Down was completed 12/13 and 7/9 days by the caregiver at the time of preliminary analysis. Preliminary efficacy will be examined upon completion of each participant’s participation and will be updated at the time of the conference. The Power Down has successfully been implemented with our first participants. As recruitment continues, interim data analysis will be conducted to examine continued acceptability, feasibility, and preliminary efficacy of the intervention. The Klingenstein Third Generation Foundation Fellowship (PI: Hartman)

Emotion regulation deficits are a hallmark of adolescent depression, and sleep greatly impacts emotion regulation. Initial data indicate acute mood benefits of slow-wave sleep deprivation (SWSD) in depressed adults, but it is unclear whether this may occur through improvement in emotion regulation. In addition, this has not been tested experimentally in adolescent depression. In this pilot study, we tested the effect of SWSD on emotion regulation in adolescents with elevated depressive symptoms. Fifteen adolescents (mean age [SD] = 17.47 [1.55] years, 12 female) completed three consecutive nights of polysomnographic sleep recording: Baseline, SWSD, and Recovery. Auditory stimulation (sounds of varying pitch, duration, and volume) suppressed slow-wave sleep (SWS) during SWSD. After Baseline and SWSD nights, the Emotional Go/No-Go task was administered the next day as a behavioural assessment of cognitive control, emotion discrimination, and emotion regulation outcomes. False Discovery Rate was used to account for multiple comparison correction. We found that, at Baseline, longer SWS duration was associated with poorer emotion discrimination (β = -0.44, p = 0.012, Q = 0.036). There was no association between other sleep stages and emotion regulation. While Emotional Go/No-Go outcomes did not significantly differ between Baseline and SWSD nights, greater attenuation in SWS significantly correlated with improvement in cognitive control (β = 0.61, p = 0.021, Q = 0.038), emotion discrimination (β = -0.44, p = 0.025, Q = 0.038), and emotion regulation (β = 0.62, p = 0.049, Q = 0.049) between nights. Findings from this pilot study tie elevated SWS to impaired emotion regulation in adolescents with depressive symptoms and suggest that targeted deprivation of SWS may improve emotion regulation in depressed adolescents with elevated SWS.

Objectives:Current approaches to objective measurement of sleep disturbances in children overlook the period prior to sleep, or the settling down time. Using machine learning techniques, we identified key features that characterize differences in activity during the settling down period that differentiate children with sensory sensitivities to tactile input (SS) and children without sensitivities (NSS). Methods:Actigraphy data were collected from children with SS (n = 17) and children with NSS (n = 18) over 2 weeks (a total of 430 evenings). The settling down period, indicated using caregiver report and actigraphy indices, was isolated each evening and seven features (mean magnitude, maximum magnitude, kurtosis, skewness, Shannon entropy, standard deviation, and interquartile range) were extracted. 10-fold cross-validation with random forests were used to determine accuracy, sensitivity, and specificity of differentiating groups. Results:We could accurately differentiate groups (accuracy = 83%, specificity = 83%, sensitivity = 84%). Feature importance maps identify that children with SS have higher maximum bouts of activity (U = -2.23, p = 0.026) during the settling down time and a higher variance in activity for the children with SS (e.g., interquartile range, Shannon entropy) that sets them apart from their peers. Conclusion:We present a novel use of machine learning techniques that successfully uncovered differentiating features within the settling down period for our groups. These differences have been difficult to capture using standard sleep and rest-activity metrics. Our data suggests that activity during the settling down period may be a unique target for future research for children with SS.

Sleep plays a vital role in brain development during adolescence. However, the relationship between sleep architecture and white matter remains poorly understood. Additionally, little is known about how puberty influences these associations, despite adolescence being a period of significant sleep and white matter changes. We investigated the association between sleep architecture and white matter integrity during early adolescence and examined whether these relationships vary with puberty. The uncinate fasciculus was selected due to its association with sleep and its role in emotion and memory, making it particularly relevant for studying sleep-related effects. A total of 455 adolescents (Female=217; Age range=10.75-13.41 years; Mean age[SD]=11.97[0.65]years) from the Adolescent Brain Cognitive Development (ABCD) study with good quality diffusion MRI and Fitbit wrist monitoring were included. Sleep architecture (light, deep, and REM) was estimated from Fitbit data. Neurite Orientation Dispersion and Density Imaging (NODDI) metrics were derived from diffusion MRI: Neurite Density Index (NDI) to measure neurite density and Orientation Dispersion Index (ODI) to assess neurite organization. Linear regressions tested cross-sectional associations between sleep stages and white matter. Pubertal status (136 prepubertal vs. 319 undergoing puberty) was included as an interaction. Age and sex at birth were covariates. Longer duration of REM sleep was associated with greater uncinate fasciculus NDI (β=0.27,P=.003). Shorter duration of light (β=-0.21,P=.008) and deep (β=-0.14,P=.005) sleep were associated with greater uncinate fasciculus ODI. While the effect of deep sleep was consistent across puberty groups, prepubertal adolescents showed stronger effects of REM sleep on NDI (β=0.24,P=.006) and light sleep on ODI (β=-0.20,P=.017) relative to those already undergoing puberty. This analysis highlights the role of sleep in white matter development during early adolescence, with stage-specific and puberty-dependent associations between sleep architecture and white matter. REM sleep enhance neurite density (↑NDI), potentially supporting axonal growth. In contrast, light and deep sleep improve white matter organization (↓ODI), crucial for efficient brain network connectivity. These findings provide new insights into sleep’s associations with white matter structure and imply differential importance of specific sleep physiological features for white matter development by developmental stage. Brain & Behavior Research Foundation (PI: Lima Santos).

Purpose Sleep is vital for brain development. Animal models have suggested that insufficient sleep affects axons and dendrites (known as neurites). However, the effects of insufficient sleep on neurites during brain development in humans remain understudied. Deriving neurite density index and orientation dispersion index (ODI) in a large sample (N = 1,016; 47.44% girls), we aimed to identify the effects of insufficient sleep on white matter development between late childhood (mean age [standard deviation] = 9.96 [0.62] years) and early adolescence (mean age [standard deviation] = 11.94 [0.64] years). Methods Longitudinal Latent Class Analysis was used to derive longitudinal classes based on sleep duration from the Sleep Disturbance Scale for Children. The Child Behavior Checklist characterized behavioral (internalizing: anxious/depressed, withdrawn/depressed, somatic; externalizing: social, thought, attention, rule-breaking, and aggressive) problems. Regression analyses evaluated the effects of sleep classes on neurite density index, ODI, and standard tensor-based metrics (Fractional Anisotropy) changes over time, the focal or widespread effects along the tracts, and whether these effects were associated with behavioral problems. Results Insufficient (<9 hours; N = 569) and sufficient sleep (>9 hours; N = 447) groups were identified. Insufficient sleep was associated with worsening fiber coherence (greater ODI) in most tracts, including cingulum bundle (F(1,982) = 9.22, p = .002, Q = 0.009), forceps minor (F(1,982) = 5.30, p = .021, Q = 0.026), and superior longitudinal fasciculus (F(1,982) = 7.41, p = .007, Q = 0.015). These effects were focal, particularly in the frontal portions of the tracts. No other metric was affected (p > .050). In addition, greater ODI in the cingulum bundle was associated with more anxious/depressed problems (β = 0.10, p = .012, Q = 0.036). Discussion Our findings suggest that insufficient sleep during this sensitive period affects white matter development, which in turn affects internalizing problems. Our findings support the importance of promoting sufficient sleep during early adolescence.

Research has sought to understand insomnia through identification of subtypes, yet age of onset has received limited focused empirical attention. This investigation aimed to detect clinically distinct age of insomnia onset subgroups utilising latent profile analysis (LPA). Participants were 618 adults, aged 18-71 years (M = 28.94, SD = 11.06), with insomnia. Participants completed a survey assessing insomnia natural history and causal attributions; sleep disturbance and impairment; pre-sleep arousal and threat monitoring; stress, mental and physical health; and social functioning. LPA was performed on age of insomnia onset. Binary logistic regression analyses were performed to evaluate associations between clinical measures and early versus late onset insomnia with statistical adjustment for chronological age and sex. Results showed a two-class model (Class 1: n = 547, 88.5%, Monset = 19.21 years, range = 0-34 years; Class 2: n = 71, 11.5%, Monset = 43.49 years, range = 35-68 years) was optimal for forming insomnia age of onset subgroups. Bodily arousal and developmental (e.g., childhood experiences, traumatic events) contributors to insomnia onset, greater overall and cognitive pre-sleep arousal, later bedtime and rise time, greater depressive symptoms, and endorsement of lifetime major depressive disorder, migraine, and arthritis were significant indicators of early onset insomnia subgroup membership. Hormonal contributors (e.g., ageing, menopause) to insomnia onset and maintenance, and more positive global mental health were significant indicators of late onset insomnia subgroup membership. Findings suggest the relevance of mindfulness-based, acceptance-based, and trauma-focused adaptations of cognitive-behavioural therapy for early onset insomnia, and management of ageing-related hormonal changes for late onset insomnia.

Circadian modulation of mood has been documented across several age groups and contexts but are less well documented in early/middle adolescents. We present novel findings on the trajectory of mood and alertness among early adolescents during a controlled 36-hour ultradian protocol. Adolescents (N=55, ages 13–15.9 yrs) participated in a 36-hour ultradian protocol consisting of 2-hour cycles of 80-minutes of wakefulness followed by a 40-minute sleep opportunity beginning at 9:00 AM. Mood and alertness were assessed during each cycle using 100-point visual analog scales (higher scores: more positive mood, higher alertness). Participants were time isolated and in dim light conditions (< 10 lux) for the first 27.5 hours. Circadian and homeostatic trends were analyzed via Cosinor mixed effect modeling. Mood fluctuated significantly in a circadian pattern (B:cos=-2.21, [95% CI=-2.85, -1.56]; B:sin=2.43 [1.77, 3.10]; p < 0.001), with an amplitude of 3 units (95% CI: 2.6–3.9). The mean mood level (mesor) was 74.44 (95% CI: 70.9–77.7), with peaks (acrophase) occurring 8.81 hours into the protocol (5:49 PM; nadir at 5:49 AM). Mood declined slightly over the course of the 36-hour protocol at an average rate of 0.11 points per hour. Alertness also showed significant fluctuations (B:cos=-7.23, [95% CI=-8.41, -6.06]; B:sin=8.05 [6.83, 9.27]; p < 0.001), with an amplitude of 10.83 units (95% CI: 9.7–12.0) and a mean level (mesor) of 56 (CI: 51.5–60.0). Peaks occurred 8.79 hours into the protocol (5:47 PM; nadir at 5:47 AM), and alertness declined steadily at an average of 0.27 points per hour. During an ultradian protocol, self-reported mood and alertness indicated a clear circadian pattern that declines across the night and increases the following day, though with some evidence of accrual of a sleep debt across the 36-hour protocol. Our finding of a ~5:48 PM peak in mood and alertness aligns with previous naturalistic studies of daily rhythms in positive mood that have reported peaks around in mid-afternoon/early evening. Thus, daily rhythms in mood may have an endogenous origin rather than being driven solely by sociocultural factors, such as the end of the school/work day. P50 DA046346

Adolescence is thought to involve dramatic changes in circadian rhythms and homeostatic sleep drive, but data demonstrating such changes in carefully-controlled laboratory studies remain sparse. We used an ultradian sleep-wake protocol to examine circadian rhythmicity in melatonin, core body temperature (CBT), and task performance in relation to age among early/middle adolescents. Participants (N=54, ages 13.0–15.9) completed 2 weeks of actigraphy followed by 60-hour lab visit including a baseline night of polysomnography (PSG), followed by a 36-hour ultradian protocol in dim light (2-hour cycles of 80 minutes awake and a 40-minute sleep opportunity with PSG), and a recovery PSG night. Core body temperature (CBT) was measured continuously and salivary melatonin was sampled every 30–60 minutes. During each ultradian cycle, participants completed the psychomotor vigilance task (PVT) to index sustained attention (outcome: lapses) and the reward anti-saccade (RAS) task to assess inhibition of a prepotent response (outcome: accuracy). Circadian trends in performance and total sleep time (TST) during 40-minute sleep opportunities were examined with mixed effects cosinor models. Peak melatonin time was estimated using spline models. Associations between age and acrophase across outcomes were examined with bivariate correlations; associations between sleep, physiologic, and performance rhythms were examined with age-adjusted partial correlations. Significant circadian trends (all p’s< 0.001) were observed in TST and PVT/RAS performance. In bivariate correlations, age was significantly associated with habitual sleep duration (r=-0.24, p=0.05), DLMO (r=0.37, p=0.018), and peak melatonin time (r=0.33, p=0.05), but not with habitual midsleep timing, acrophase of performance rhythms, or TST acrophase. Age-adjusted partial correlations indicated that DLMO was associated with peak melatonin time (r=0.59, p< 0.001), and marginally associated with CBT nadir (r=0.36, p=0.065) and PVT acrophase (r=0.33, p=0.096). CBT nadir was associated with habitual midsleep (r=0.49, p=0.005), TST acrophase (r=0.73, p< 0.001), and peak melatonin time (r=0.38, p=0.045). In this sample of adolescents, age was associated with some, but not all sleep and circadian rhythm measures. Although strong circadian rhythms were observed in physiological and performance measures, the timing of these rhythms were not consistently associated with each other. Changes in circadian rhythms during adolescence may vary among different domains. P50 DA046346

Objectives:Adolescence is marked by significant changes in sleep physiology, which provides crucial insights into brain development and health outcomes. However, sleep has been difficult to measure on a large scale due to the limitations of polysomnography. Emerging portable devices, like the Dreem3 headband (Dreem3), offer a scalable and accessible alternative, with validation studies in adults showing strong feasibility and accuracy, but validation data for adolescents remain limited. This pilot study evaluated the acceptability and feasibility of the Dreem3, and day-to-day stability of Dreem3-derived sleep estimates, in adolescents and young adults. Methods:Eighty-one participants completed 3 consecutive nights of sleep recordings with the Dreem3: 32 Adolescents (9-17yrs) and 49 Young Adults (18-26yrs). We evaluated acceptability with a user experience survey and determined feasibility based on completion and quality of sleep recordings. We used intraclass correlation coefficients to assess between-night stability of sleep macro-architecture and micro-architecture estimates. Results:Dreem3 was similarly acceptable for Adolescents and Young Adults (p-values>0.1) on most user experience survey outcomes except that the Young Adult group reported poorer sleep quality on Dreem3 nights relative to the Adolescent group (p<0.05). The number of recording nights (total and good quality) between-night stability of sleep architecture did not differ between groups (p-values>0.05). Conclusion:Our pilot data indicates that Dreem3, is well-accepted and feasible in young people, and provide moderately stable sleep data across nights. These findings open the door to large-scale, at-home sleep studies with portable sleep recording devices in adolescents.