Heterogeneous Neurophenotypes of Adolescent Sleep Insufficiency Stratify Natural Short Sleepers from Comorbidity or Environment Driven Insufficiency

Heterogeneous Neurophenotypes of Adolescent Sleep Insufficiency Stratify Natural Short Sleepers from Comorbidity or Environment Driven Insufficiency | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Heterogeneous Neurophenotypes of Adolescent Sleep Insufficiency Stratify Natural Short Sleepers from Comorbidity or Environment Driven Insufficiency Dan Wu, Yiwei Chen, Mingyang Li, Zhiyong Zhao, Xinyi Xu, Yongquan Huang, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6851158/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Mar, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Insufficient sleep has become an increasing public health issue in adolescents, which is associated with complex social-ecological factors and a wide range of neurodevelopmental outcomes. The highly heterogeneous determinants and outcomes of insufficient sleep impose challenges in effective interventions. To understand such heterogeneity in terms of its impact on the brain, we employed a data-driven ​Subtype and Stage Inference (SuStaIn)​ model to classify the spatiotemporal trajectories of MRI-based neurobiology in 3,266 adolescents from the ABCD study (853 of them had sleep duration less than 8 hours according to Fitbit measurements). We identified three distinct subtypes with reduced cortical thickness, starting from the postcentral cortex, pericalcarine cortex, and entorhinal cortex, respectively. These subtypes diverged significantly in sleep-related social-ecological factors. The postcentral-originated subtype mirrored healthy controls in sleep behavior and sleep environment and showed no psychiatric comorbidities, which aligned phenotypically with natural short sleepers. Notably, this subtype displayed significantly advanced brain age, suggesting advanced neurodevelopment that explained the reduced sleep demand, and polygenetic score analysis revealed a genetic predisposition for short sleep in these adolescents. The pericalcarine-originated subtype displayed environmentally driven sleep insufficiency (e.g., light/noise pollution) where sleep duration mediated environmental effects on pericalcarine cortical thinning. The entorhinal subtype showed elevated psychiatric risk, younger brain age, and spatial correlations with psychosis-related neurotransmitter systems. This work deciphered heterogeneous impacts of insufficient sleep on the brain and the relation with biological, mental, and environmental determinants, offering a framework to guide stratified prevention and intervention strategies. Health sciences/Health care/Paediatrics/Paediatric research Health sciences/Health care/Disease prevention/Lifestyle modification Health sciences/Biomarkers/Prognostic markers Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Insufficient sleep is prevalent in modern society, affecting 1/3 of the population 1 . Particularly, sleep problems in adolescents have become a major concern due to changes in the social environment, which are complicated by rapid neurobiological development 2 . According to previous studies, adolescent sleep of less than 8 hours is defined as insufficient sleep, affecting 16% of sixth graders 3 . A good night's sleep is hard to have, as it is affected by various factors 4 , including physiological changes of puberty 5 , academic stress load 6 , mental health 7 , and sleeping environments 8 . Insufficient sleep also causes major social and health consequences, including both physical 9,10 and mental fatigue 11 , negatively influencing social life 12 and school performance 13,14 . A recent study reported that sleep disturbances emerged as the most influential predictor of mental health risk 15 . The multifaceted and heterogeneous determinants and outcomes of insufficient sleep hinder our understanding of sleep in children and adolescents 16 . Previous studies used the social-ecological framework to understand the etiology of sleep problems. Under such a framework, causes of insufficient sleep were divided into three layers, namely, sleep-related factors, personal conditions (mental or physical discomfort), and environmental factors. In practice, some children may be biologically disposed to short sleep duration without compromised development 17 , while others sleep less because of external disturbance 18 , psychiatric comorbidity 19 , or other conditions, that require entirely different interventions. To settle such a problem, a reliable, unbiased, and informative biomarker becomes the key for stratification. To capture the different sleep profiles, previous studies focused on various objective and subjective sleep measures using latent profile analysis 20,21 . Other studies of sleep genetics 22 and proteomics 23 found genes and proteins for sleep regulation. On the other hand, as insufficient sleep has major impacts on brain development 24 , an in-depth understanding of neurobiology needs to be further addressed. In recent years, MRI-based neuroimaging approaches have been used frequently to unravel the underlying neurological mechanisms of sleep problems. Cross-sectional studies found that shorter sleep time is correlated with lower cortical brain volume for adolescents 25,26 . A longitudinal study confirmed that shorter objective sleep time is correlated with a thinner cortex 27 . A review summarized that insufficient sleep strongly correlates with neuroanatomical structures and functions of adolescent brain development, but due to the heterogeneity of insufficient sleep, a precise conclusion cannot be drawn 28 . Other studies also found significant correlations between subjective 29 or objective 30 sleep measures and distinctive brain phenotypes, further confirming the presence of insufficient sleep brain subtypes. Therefore, although at the group level, the current findings pointed to cortical thinning and reduced cortical volumes in the adolescent population with insufficient sleep, the substantial inter-personal differences among individuals 16 demand better stratification. The rich anatomical and functional information and the unbiased nature of neuroimaging make it an ideal neurobiological marker for the stratification of sleep disorders. For instance, previous studies found different insomnia subtypes in adults with distinctive cortical topography and structural connectivity 31 . It is not yet known whether individuals with insufficient sleep (i.e., short sleep duration) exhibit subgroups with distinct brain patterns, especially in the adolescent population, and whether the imaging phenotypes possess explicit social-ecological characteristics that can help to inform effective interventional strategies. In this study, we aim to decipher the social-ecological heterogeneity of insufficient sleep in adolescents based on brain MRI, including the biological, mental, and environmental factors. We adopted the Subtype and Stage Inference (SuStaIn) 32 method to identify insufficient sleep subtypes based on cortical thickness (CT) using the Adolescent Brain Cognitive Development (ABCD) study with MRI and Fitbit-based sleep measurements. An optimal number of three subtypes was found using the Bayesian model selection and was validated using the longitudinal data from the ABCD dataset and a house-collected independent dataset. To understand the socio-ecological determinants underlying these subtypes, we comprehensively examined pediatric sleep health, physical or mental conditions, and family and neighborhood factors among these subtypes. Lastly, we compared the brain ages, polygenic score (PGS), associations with psychiatric disorder, and neurotransmitter profiles among these subtypes. Figure 1 provides an overview of the present study. Results Brain morphology-based subtyping of insufficient sleep We used objective sleep duration measured by Fitbit to characterize insufficient sleep. According to guidelines 33,34 , adolescents with objectively measured sleep of less than 8 hours were defined as insufficient sleepers. We also compared other objective sleep measurements between the insufficient and sufficient sleep groups. The insufficient subtype had significantly higher sleep efficiency (P Bonferroni < 1x10 − 5 ), fewer average wake counts per night (P Bonferroni < 1x10 − 5 ), and similar sleep latency (P Bonferroni = 0.71). We first identified the brain patterns most susceptible to sleep deprivation by comparing the insufficient and sufficient sleep groups. Using the Desikan-Killiany (DK) Atlas as a template, we identified the region of interest (ROI) with significantly altered cortical thickness (CT) across the whole-brain after controlling for random factors of age, sex, ethnicity, BMI and site, using subjects from the ABCD second-year follow-up ( N insufficient sleep = 853, N sufficient sleep = 2,413). Figure 2 a demonstrated that sleep had a profound effect on the developing adolescent brain. 29 of 68 brain regions showed reduced CT in the insufficient sleep group compared to the controls, with the left lateral orbitofrontal cortex showing the most reduced CT with Cohen’s d of 0.200 (P Bonferroni = 7.95x10 − 5 ). We then used the SuStaIn model to identify distinct longitudinal trajectories of CT reduction with three z-score cutoffs: 1, 1.5, and 2. A three-cluster model stood out with minimum cross-validation information criterion (Supplementary Fig. S2 a) and maximal log-likelihood (Supplementary Fig. S2 b) under ten-fold validation. This resulted in three subtypes with distinctive topography, i.e., subtype 1 showed CT reduction across the whole whole-brain (n = 197, 23.09% of the total), subtype 2 showed a lower CT in the occipital lobe (n = 396, 46.42% of the total), and subtype 3 showed a lower CT in the entorhinal cortex (n = 260, 30.48% of the total). To test the stability of the results, other z-score cutoffs (1, 2, and 3; 2, 3, and 4) for SuStaIn were also tested (Supplementary Fig. S4). Sleep deficit-induced brain changes also showed different progressive trajectories among the subtypes. Subtype 1 exhibited initial change in the postcentral cortex, which then progressed to the supramarginal precuneus and inferior parietal area, namely, the postcentral-originated subtype. Subtype 2 showed initial effects in the pericalcarine area, which then progressed to the lingual area, namely, the pericalcarine subtype. Subtype 3 first demonstrated alteration in the entorhinal cortex, then progressed to the inferior temporal and precentral cortex, namely, the entorhinal subtype. The SuStaIn model assigned each insufficient sleep adolescent with the most likely subtype and stage label. The subjects assigned to ‘stage 0’ of the progression stage were noted as ‘pre-effect’, while others were noted as ‘post-effect’. The probability-based distribution of post-effect individuals in the three subtypes was depicted in Fig. 2 c. Figure 2 d illustrated the different stages in the three subtypes of sleep deficit, represented as the mean z-value images, i.e., reduction of CT in the ‘post-effect’ phase compared to the controls. Longitudinal change of brain morphology in the three subtypes To measure the change of brain CT and compare it with pseudo staging from SuStaIn, we obtained ΔCT by subtracting the CT values between two rounds of longitudinal MRI scans while regressing the age gap between scans and other covariates. We measured the variability of ΔCT by log-transformed Euclidean distances within groups, and found the insufficient sleep group had higher within-group variability under multiple atlases (DK and Destrieux atlas, Fig. 3 a), supporting the higher heterogeneity and the necessity for subtyping. After subtyping, we found the within-group variability of ΔCT for the postcentral-originated subtype no longer showed significant difference compared with the sufficient sleep group, indicating the similarity of this subgroup to the controls (Fig. 3 b). Fig. 3c demonstrated that the postcentral subtype showed much greater ΔCT in the parietal lobe, while the pericalcarine subtype showed greater ΔCT in the occipital lobe, and the entorhinal subtype showed greater ΔCT in the temporal cortex, which were in accordance with the “pseudo progression” obtained from SuStaIn model (Fig. 3d). Characterization of subtypes by social-ecological profiles Previous studies used socio-ecological and systems frameworks to understand the determinants of sleep patterns and problems in children and adolescents 35–37 . To take the layers of socio-ecological factors (Fig. 4a) into full account, we compared the insufficient sleep subtypes with the control group in terms of sleep factors, physical or mental conditions, and neighborhood environment (Fig. 4b), while controlling for confounders including age, sex, BMI, site, race, ethnicity, socioeconomic status. Interestingly, no significant differences were found between the postcentral subtype and the control for any of the socio-ecological factors. Therefore, we interpreted this subtype as natural short sleepers who although sleep relatively less, do not show compromised development, and thus may not require any intervention. We further revealed this group exhibited advanced brain age that may explain the lower CT and increased ΔCT in this group. Comparing the polygenetic risk score (PRS) based on GWAS data of short sleep duration, this group also showed a short sleep duration genetic predisposition. The pericalcarine subtype was found under adverse environmental conditions (brighter night light, louder noise at night, less neighborhood income, lower neighborhood education, and more serious neighborhood disadvantage), and worse physical activities, leading to sleep problems (initiating and maintaining sleep and weekly sleep loss). Therefore, we labeled this subtype as environment-induced sleep deficits. We further performed mediation analysis to investigate the relationship between night light, sleep duration, and pericalcarine in this group. The entorhinal subtype was mostly affected in pediatric sleep domains, including initiating and maintaining sleep and weekly sleep loss and weekly sleep loss. By examing the linear correlations between SuStaIn stages and residualized CBCL score, we found higher staging in the entorhinal subtype were associated higher externalizing problems ( r = 0.394, P Bonferroni = 0.008, Bonferroni corrected; Fig. 5f), higher internalizing problems ( r = 0.319, P Bonferroni = 0.032, Bonferroni corrected; Fig. 5g), and overall problems (r = 0.30, P Bonferroni = 0.045, Bonferroni corrected; Fig. 5h). Given that this subtype showed higher degree of psychiatric comorbidity compared to the other subtypes, we characterized this group as comorbidity-related. The overall socio-ecological profiles of the subtypes were summarized in Fig. 4d. Brain age differences among subtypes During adolescence, the need for sleep duration strongly declines with brain maturation 38 . As brain age derived from neuroimaging has been widely used as a biomarker for brain aging and maturation 39 , we tested whether the brain age differed among the subtypes, although the chronological age showed no significant difference (Fig. S7). Compared with the control, the postcentral subtype (aka. short sleepers) showed higher (P Bonferroni < 0.001, Bonferroni corrected; Fig. 5 a) and the entorhinal (aka. comorbid related) showed lower (P Bonferroni = 0.022, Bonferroni corrected; Fig. 5 a) brain age. This result indicated more advanced development in the postcentral subtype (the short-sleepers), which well explained the relatively low need for long sleep duration. The lower brain age in the entorhinal subtype suggested potentially delayed development, potentially due to the psychiatric comorbidity. Polygenetic risk score of short sleep among subtypes Previous studies found a significant genetic basis for short sleepers 17,40 . Hassan et.al 40 found 27 significant loci correlated with short sleepers in adults (subjective sleep < 7 hours), and were further validated with accelerometer-measured objective sleep. We calculated the short sleep duration polygenetic risk score (PRS, details see Method). The postcentral subtype had a significantly higher PRS value compared with the control (P = 0.0042; Fig. 5 b), which indicated a genetic predisposition for short sleep duration. The other insufficient sleep subtypes showed no significant differences in PRS compared with the control. Mediating effects of sleep on the environment and the brain To further understand the potential environmental effect on the pericalcarine subtype, we used mediation analysis to determine whether the associations between night light pollution and CT changes in the pericalcarine cortex could be explained by sleep duration. In the pericalcarine subtype, sleep duration mediated the association between nightlight and pericalcarine CT (path C: -0.061, 95% CI [-0.112, -0.010], p = 0.022; path a*b: -0.021, 95% CI [-0.037, -0.009], p < 0.001; Fig. 5 d). The mediation effects were not significant for the postcentral (path C: -0.056, 95% CI [-0.123, 0.010], p = 0.094; path a*b: -0.002, 95% CI [-0.008, 0.003], p = 0.486; Fig. 5 c) or entorhinal (path C: -0.027, 95% CI [-0.091, 0.037], p = 0.406; path a*b: 0.004, 95% CI [-0.002, 0.012], p = 0.166; Fig. 5 e) subtypes. Neurotransmitter basis of insufficient sleep subtypes Studies using PET and SPECT measured a range of receptors and transporters highly correlated with psychiatric disorders. This enabled us to correlate the spatial patterns of CT in each subtype and the density maps of neurotransmitter and transporter systems (5-hydroxytryptamine: 5HT1a, 5HT1b, 5HT2, and 5HT4; dopamine: D1, D2, and DAT; GABA: GABAa). We found the postcentral subtype (Fig. 5 g) was significantly correlated with D2, DAT that involved in dopaminergic signaling (r=-0.428, 0.377, respectively, all P < 0.001). The pericalcarine subtype (Fig. 5 i) was significantly correlated with a serotonin receptor 5HT1a (r = 0.33, p = 0.001). The entorhinal subtype (Fig. 5 g) was negatively correlated with 5HT1a (r=-0.613), and positively correlated with D2 and GABAa (r = 0.507, 0.511 respectively) (Fig. 5 h). Validation of the subtyping results To test the robustness of subtyping results against different definitions of sleep insufficiency, we applied the SuStaIn model to other insufficient sleep cutoffs (7.5 and 8.5 hours) and compared the sequences of biomarker progression with the primary results using the 8 hours cut off (Fig. S3). The SuStaIn trajectories of both replication cutoffs showed high consistency with the original cutoff, with the correlation coefficients ( r ) of 0.785, 0.811, and 0.689 for the three subtypes between 8 hours and 7.5 hours cutoffs, respectively (P one−tailed < 0.001, FDR corrected, Fig. S3e) and r of 0.857, 0.921, and 0.811 for the three subtypes between 8 hours and 8.5 hours cutoffs (P one−tailed < 0.001, FDR corrected; Fig. S3f). We further validated the results with subjective sleep cutoff (< 8 hours of sleep), which also showed a high consistency (Fig. S6). We then validated the results on our private dataset (Shanghai Sleep Birth Cohort, SSBC) using the identical procedure. The SuStaIn subtypes found in the SSBC dataset showed high consistency with the discovery set of the ABCD dataset, e.g., subtypes showed starting epicenters in the postcentral cortex (Fig. 6 a), pericalcarine cortex (Fig. 6 b), and entorhinal cortex (Fig. 6 c). For the SSBC dataset, the postcentral subtype was found with similar pediatric sleep presentations as the short sleeper found in ABCD; while the pericalcarine and entorhinal subtype were found of worse pediatric sleep presentations compared with the control group (Fig. 6 d). The stage progression in the entorhinal subtype was found to be significantly correlated with externalizing problem and the overall problem scores (Fig. 6 e-g). Such results were also highly in accordance with results from the ABCD study. However, no significant correlation was found between internalizing problems and SuStaIn stage. Discussion Insufficient sleep has a long-term and heterogeneous impact on the physical and mental health of adolescents 41 , but little is known about its neurobiological mechanisms 42 , especially its progressive and accumulative effect on brain development 24,43 . Previous studies and guidelines for adolescent sleep recommend more than 8 hours of sleep to maximize health and well-being 33,34 . Although sleep duration recommendations inform public policies, medical guidelines and interventions, individual sleep requirements 44,45 and the cause of insufficient sleep duration 35 vary. In this work, we used a data-driven approach (SuStaIn) on a large adolescent sample with both neuroimaging and Fitbit measurement, and identified three subtypes of insufficient sleep based on their cortical thickness, which were associated with distinct socio-ecological factors. Compared with the controls, the pericalcarine subtype was under adverse environmental conditions, along with worse sleep qualities. The entorhinal subtype had a high chance of comorbidity with mental disorder-related behavioral problems, worse sleep qualities, and younger brain age, and was significantly associated with. The postcentral subtype did not show impaired sleep qualities or any comorbidity and showed more advanced development in terms of brain age. Furthermore, the robustness of the three subtypes and their staging was validated with longitudinal data in ABCD, different sleep duration cutoffs, and an independent dataset. These findings emphasize the clinical potential of using brain morphological features for stratification of sleep deficits, as the distinct socio-ecological determinants of different subtypes indicated entirely different interventions. One interesting finding in this study is that we identified a subtype of natural short sleepers. It is known that biological demand for total sleep duration is highly personalized 17 . Previous studies also pointed out that there is a group of natural short sleepers who are satisfied with extremely less sleep time, based on genetics, sleep, and other mental or physical measurements. 17,46 In comparison with the control group, despite significantly lower sleeping hours (7.46 +/- 0.51 hours), this postcentral subtype had similar presentations in all sleep-related socio-ecological factors and no negative psychiatric outcomes. Moreover, the postcentral subtype had a significantly higher PRS score for short sleep duration genes, indicating a genetic predisposition for short sleep duration. We also found the shorter sleepers had relatively fewer night wakes in both the ABCD and SSBC datasets (p = 0.051 for the ABCD study and p = 0.161 for the Shanghai dataset; marginally significant), indicating higher sleep efficiency in these adolescents. In the brain development perspective, we found the parietal cortex in this subtype was thinner and reduced faster longitudinally than the controls. Previous studies have found significant correlations between parietal cortex and physical activities in structural and functional domains 47–49 . Compared with other sleep subtypes, the natural short sleepers had a relatively higher physical activity, which may explain the morphological change of the parietal cortex and help to mitigate the negative effects of sleep loss. This speculation was supported by the “brain age”, as the short sleepers were found to have a more advanced brain age compared with the controls, which indicated a significantly higher brain maturity and fast development. As the normative sleep duration decreases during adolescence 50,51 , the more advanced development in this subgroup suggested a natural need for less sleep time. Therefore, it is important to identify this population of nature short sleepers to avoid unnecessary concerns or interventional plans. For the other two groups, we identified various external and internal stimulants that may induce sleep deficits and further impair brain development. Sleep environment, including ambient light, noise, and neighborhood depravity, significantly affects adolescent sleep, leading to negative mental and physical consequences 52,53 . A missing link in this causal relationship is how negative sleep environments change the brain. Preclinical models demonstrated that sleep deprivation during adolescence disrupts synaptic pruning 54 —a proposed mechanism for cortical thinning 55 . Among the three SuStaIn subtypes, the pericalcarine subtype was found with worse sleep environmental conditions. As part of the visual network, the pericalcarine cortex exhibits heightened sensitivity to sleep-environment factors. Previous studies found that REM sleep loss (associated with light fragmentation) correlated with reduced default mode network-visual cortex connectivity 56 . We further found that sleep duration mediated ambient light and pericalcarine cortical thickness. These results support the light pollution mechanism proposed by LeGates et al. 57 that light pollution disrupts circadian rhythms and sleep via ipRGCs, which may contribute to brain morphological changes by chronically impairing neuroplasticity. Besides the external stimulants, many psychiatric disorders are related to sleep deficits 7,11,19,58 . A number of studies have demonstrated that the emergence of psychiatric disorders is partially related to morphological changes caused by abnormal synaptic pruning and myelination 59–61 . For the entorhinal subtype, we found a strong association with psychiatric comorbidity, including both internalizing and externalizing problems. The entorhinal cortex plays an important role in memory processing, mediating cortical information from the hippocampus 62 . Acting as a “gateway” between the hippocampus and neocortex 63 , abnormal thinning of the entorhinal cortex may reflect the impaired cognitive functions 64 due to the comorbidity. During adolescence, developmental changes in neurochemical systems, including dopaminergic and GABA, reform adolescent motivation behaviors, leading to potential internalizing and externalizing disorder appearances 60 . Further analysis revealed that the entorhinal subtype had a much stronger spatial correlation with these psychiatric-related neurotransmitters. The entorhinal subtype was also found to be significantly younger in brain age, suggesting a delay in brain development and indicating potential correlations with psychiatric disorders 65–67 . According to the layers of sleep socio-ecological factors 35,36 , short sleep duration is associated with environmental factors, physical or mental conditions, and pediatric sleep health issues. The three SuStaIn subtypes are distinctively related to different aspects of sleep socio-ecological factors, and the findings were highly robust, as validated with ABCD longitudinal data and an independent dataset. Several issues should be taken into account for our findings. Firstly, the SuStaIn models pseudo-longitudinal sequences via cross-sectional data, which do not directly reflect the actual progression. Although longitudinal data supported SuStaIn trajectories (ΔCT patterns reflected the SuStaIn subtype trajectories), future work needs to verify the sleep-induced brain alteration trajectories. Secondly, as sleep health is a multidimensional construct, future studies can incorporate more sleep-related variables, including sleep timing and sleep efficiency, to have a more comprehensive stratification of sleep. Moreover, our present research focused on the effects of insufficient sleep on cortical thickness; further studies could incorporate other imaging features and genomic data. In addition, while clustering or subtyping can facilitate stratified interventions, individual differences within clusters still exist. Therefore, dimensional approaches 68 to personalization may further address the individual sleep variations. Thirdly, although the objective sleep measurements in the ABCD study (with Fitbit) and the SSBC dataset (with actigraphy) were measured differently, both Fitbit and actigraphy are accurate measurement tools for sleep duration 69 . Validation experiments with both subjective and objective measurements of sleep duration and cutoffs also support the robustness of insufficient subtype results. Finally, the subjects of PET or SPECT neurochemical density maps were all adults, while our study subjects were adolescents. Further lifespan works could fill this gap towards a better understanding of the adolescent population. In summary, we used a data-driven method to disentangle the heterogeneity of the impact of insufficient sleep on the brain, identifying three subtypes with distinct and replicable neuroimaging patterns. This was followed by a comprehensive analysis that revealed the group differences in biological, mental, and environmental determinants. Such results may offer a comprehensive understanding of insufficient sleep on adolescents, calling for stratified attention and interventions towards this social problem. Method Data preparation In the second-year follow-up of the ABCD study, 4,134 adolescents with Fitbit worn on weekdays and weekends and with at least seven nights of recorded sleep were included 70 . Note that the baseline ABCD data do not have Fitbit measurements, and thus was not used for discovery. Subjects missing imaging data or covariates (age, sex, BMI, site, race, ethnicity, socioeconomic status) were excluded, leaving 3,266 individuals for formal analysis. The participant inclusion flowchart was provided in Supplementary Fig. S1 . To derive an average weekday-weekend sleep profile, we calculated the sleep hours weighted according to the number of days [(∑weekday sleep duration) + (∑weekend sleep duration)]/(weekday day counts + weekend day counts) and winsorized 21 . Participants were divided into sufficient or insufficient sleep groups based on a cutoff of 8 hours 3,50 of average sleep based on Fitbit, resulting in 853 adolescents with insufficient sleep and 2,413 with sufficient sleep (age 12.03 ± 0.63 years, 56.5% male for insufficient sleep; age 11.87 ± 0.64 years, 50.6% male for sufficient sleep). We also tested other choices of 7.5 hours and 8.5 hours as cutoffs (Supplementary Fig. S2 ). The demographic characteristics are summarized in Supplementary Table S1 . A private dataset of the Shanghai Sleep Birth Cohort 71 (SSBC dataset) was used for the replication test. This dataset included 52 participants with insufficient sleep and 63 with sufficient sleep based on actigraphy (age 10.33 ± 0.27, 51.9% male for insufficient sleep; age 10.33 ± 0.25 years, 38.7% male for sufficient sleep). As subjects from the SSBC dataset were younger, 8.5 hours of sleep was used as the cutoff for insufficient sleep 50 . All study procedures received approval from institutional review boards at the data collecting sites, and written consent was obtained from parents, with verbal assent from the children. Image pre-processing We used the 3D T1-weighted MRI scans from each dataset. The ABCD data were preprocessed by a standard pipeline of cortical reconstruction and volumetric segmentation using FreeSurfer v7.1.1 72 ( https://surfer.nmr.mgh.harvard.edu/ ) from the ABCD study Data Analysis, Informatics, and Resources Center (DAIRC) along with standard quality control measures, reported in previous studies 73 . Aligning with the ABCD study, similar image pre-processing processes were performed for the SSBC dataset. Here we focused on the cortical thickness (CT) features that were shown to be significantly altered in previous studies. We used the Desikan-Killiany Atlas 74 template to parcellate the brain into 68 ROIs. Site-specific variations were estimated and regressed with Neuro-Combat 75 . Before comparing the CT measures between sufficient and insufficient sleep groups, propensity score matching was used to reduce bias caused by confounding variables. Specifically, covariates including age, sex, site, race, ethnicity, socioeconomic status, and BMI were controlled. Pediatric sleep measurements Sleep disturbance in ABCD participants is measured based on the Sleep Disturbance Scale for Children (SDSC) 76 with 26 questions, assessing six sleep disturbance dimensions, including Disorders of initiating and maintaining sleep, Sleep breathing disorders, Disorders of arousal, Sleep-wake transition disorders, Disorders of excessive somnolence and Sleep hyperhydrosis. For the SSBC dataset, the Children’s Sleep Habits Questionnaire (CSHQ) 77 was used to assess sleep behavior in eight aspects, includingBedtime resistance, Sleep onset delay, Sleep duration, Sleep anxiety, Night wakings, Parasomnia, Sleep breathing disorders, and Daytime sleepiness. As adolescents usually address weekday sleep debt in weekends 3 , we used weekly sleep loss ((weekly average sleep duration – workday average sleep duration) / workday count) measurement from the Munich Chronotype Questionnaire 78 for evaluation. The higher the weekly sleep loss measure, the larger the workday sleep debt repaid during freedays. Sleep environment measurements In the ABCD study, the child’s primary residential address was geocoded, and variables of the American Community Survey (5-year estimates from 2011 to 2015) were linked according to the US census tract 79 . Potentially sleep-affecting neighborhood environmental measures were selected, including night light, income, education, noise, and neighborhood disadvantage 18 . Mental and physical condition measurements For the ABCD dataset, mental health conditions of internalizing, externalizing, and overall problems were assessed with Parent Child Behavior Checklist Scores (CBCL) 80 . For the SSBC dataset, subjects’ mental conditions of internalizing, externalizing, and overall problems were assessed via the Strengths and Difficulties Questionnaire (SDQ) 81,82 . Daily physical activity is measured with step counts with Fitbit. Brain age In order to evaluate the difference in brain age between the subtypes, we used the Connectome-Based Predictive Modeling (CPM) framework 83,84 for estimating brain age based on cortical thickness. Cortical thickness from all 68 regions was standardized while regressing out sex and site. Univariate correlation analysis (Pearson’s r ) identified regions significantly associated with chronological age (all ROIs have negative correlations with age, p < 0.05, adjusted with Bonferroni correction). We used the age-correlated ROIs to train a linear regression model. With 10-fold cross-validation, the model accuracy was evaluated using Pearson’s correlation coefficient ( r = 0.17, p = 2×10 − 22 ) between predicted and chronological age. Neurotransmitter profiles Neurochemical density profiles across cortical regions were quantified through retrospective analysis of positron emission tomography (PET) and single-photon emission computed tomography (SPECT) data obtained from a cohort of healthy controls 85 . Eight receptors across three different neurotransmitter systems (5-hydroxytryptamine: 5HT1a, 5HT1b, 5HT2a, and 5TH4; dopamine: D1, D2, and DAT; GABA: GABAa; details in Supplement 2) were investigated with JuSpace 86 . These density maps were registered to the Montreal Neurological Institute (MNI) stereotaxic coordinate system and segmented into 68 anatomically defined cortical regions using the Desikan-Killiany (DK) atlas 74 . Polygenetic risk score We compute polygenetic risk score (PRS) of short sleep duration using imputed ABCD study genotype data with quality control, based on the largest GWAS data of short sleep duration (n = 446,118) from UK BioBank 40 . For sleep duration measurement in this adult dataset, participants were asked about how many hours of sleep within 24 hours, and participants with less than 7 hours of sleep were deemed as short sleepers. Among ABCD children of European ancestry, we used PRSice-2 87 to calculate individual PRSs by adding SNP alleles with weights of the SNP allele effect size estimated in a previous GWAS (short sleep population GWAS). Parameters for polygenic scoring were a clumped variation of linkage disequilibrium r2 < .10 within a 300-kb window cutoff. We calculated PRSs for each trait based on six different p-value thresholds (pTs) with pT < 5e.08 (32 SNPs), pT < 0.001 (4,009 SNP), pT < 0.005 (12,675 SNPs), pT < 0.01 (22,140 SNPs), pT < 0.05 (73,429 SNPs), pT < 0.1 (129,416 SNPs), and pT < 0.5 (422,712 SNPs), as recommended 87 . An optimal p-value threshold (pT < 0.5) was found with the highest percentage of the variance (R 2 = 0.016, p = 4.45e-05) in the outcome estimated by PRSice-2, and were further Z-score normalized. A higher PRS score indicates a genetic predisposition for short sleep duration. Subtyping of insufficient-sleep related brain morphology We employed a widely used data-driven neuroimage morphological progression model named Subtype and Stage Inference (SuStaIn) 32 , which identifies progression subtypes with an event-based model. The SuStaIn model assigns subtypes and stages to individuals based on their biomarkers. The estimated SuStaIn stages were used to capture the progression of diseases, with earlier stages meaning little biomarker changes and later stages meaning severe biomarker changes. We first regressed out confounding factors (age, sex, site, race, ethnicity, socioeconomic status, and BMI) from regions of interest (ROIs) using linear regression. Scanner-induced effects were regressed with Neuro-Combat 75 . The adjusted ROIs of the insufficient sleep group were z-scored relative to the sufficient sleep group, with a higher z-value indicating greater deviations from sufficient sleep. We selected twelve ROIs with significantly reduced CT in the insufficient sleep group as SuStaIn inputs (P FDR <0.05, Fig. 1 a, d). Based on the parameter settings outlined in previous research, we used an expectation-maximization algorithm for model initialization and used 25 random starting points to give the optimum solution. We applied 1,000,000 rounds of Markov Chain Monte Carlo (MCMC) to estimate the developmental trajectories of the identified subtypes. Three z-score cutoffs (1, 1.5, and 2) were employed to describe the amount of negative impact of sleep deficits on adolescents. We assessed subtyping solutions with 2–5 clusters, and an optimal choice of 3 was chosen according to its lower CVIC 88 and higher log-likelihood through ten-fold cross-validation (Supplementary Fig. S2 ). To test the stability of the results, other z-score cutoffs for SuStaIn were also tested in Supplementary Fig. S4. Statistical analysis After identifying SuStaIn insufficient sleep subtypes, we compared different socio-ecological factors affecting sleep among the subtypes for their pediatric sleep health, physical or mental conditions, and family and neighborhood factors using the Mann-Whitney U test and computed the effect size using Cohen’s d test. We further compared the brain age (in measurement for brain maturation) and short sleep PRS (in measurement for genetic backgrounds) among subtypes using the Mann-Whitney U test. As for the associations among night light pollution, sleep duration, and brain structure, we performed mediation analysis using sleep duration as the mediator. All variables were normalized before entering the model. Sex, age, parent education level, household income, ethnicity, BMI, and site were used as covariates. Total, direct, and indirect associations were quantified through 10,000 bootstrap iterations, with 95% bias-corrected and accelerated confidence intervals calculated. All statistical analyses were performed using the R mediation package (v4.0.0). The statistical significance threshold for these analyses was set at P Bonferroni < 0.05. To understand the subtype-specific neurochemical signatures, we performed a spatial correlation using bootstrapped Spearman correlation with 10,000 iterations between each neurochemical density matrix and cortical thickness T-statistic maps for each subtype. Longitudinal validation To test the reliability of the SuStaIn staging resultswe utilized ABCD longitudinal data at the first and the second visits, which were two years apart. We calculated the actual change of cortical thickness of individuals in comparison with the pseudo-progression of SuStaIn. The change of cortical thickness (ΔCT) was obtained by subtracting the value of the baseline year from the second-year follow-up in each ROI while regressing the age gap between the scans, which were compared between the sufficient and insufficient sleep groups. Independent validation We applied SustaIn model train via the ABCD sample on a private dataset of the Shanghai Sleep Birth Cohort. We compared the brain phenotype, pediatric sleep, and psychiatric comorbid between the identified subtypes and control. The results showed consistency. Furthermore, to test the robustness of subtyping results against different definitions of sleep insufficiency, we retrained the SustaIn on other insufficient sleep objective sleep cutoffs (7.5 and 8.5 hours). As sleep duration is also measured via structured scales, we used the Sleep Disturbance Scale for Children (SDSC) measured sleep duration as a cutoff (< 8 hours of sleep), and retrained the SustaIn model. We compared the consistency of results obtained by different cutoff criteria via linear correlation between the original and the retrained sequences of biomarker progression. Declarations Data availability The data repository houses all data generated by the Adolescent Brain Cognitive Development (ABCD) Study at https://nda.nih.gov/abcd/. Data will be available upon reasonable request to the corresponding author after institutional approval and with a signed data access agreement or with the permission of the Shanghai Children's Medical Center. Code availability Python of the SuStaIn algorithm is available on https://github.com/ucl-pond. Funding During the study, Prof. Dan Wu is supported by Ministry of Science and Technology of the People’s Republic of China (2018YFE0114600, 2021ZD0200202), National Natural Science Foundation of China (81971606, 82122032), and Science and Technology Department of Zhejiang Province (202006140, 2022C03057). Prof. Guanghai Wang is supported by the National Science and Technology Innovation 2030 Major Project of China (STI2030-Major Projects+2021ZD0204200), National Natural Science Foundation of China (82071493, 82073568), Shanghai Municipal Health Commission (2022XD056), and Innovative Research Team of High-level Local Universities in Shanghai (SHSMU-ZDCX20211100). Competing interests The authors declare no competing interests. References Altman NG, Izci-Balserak B, Schopfer E, et al. Sleep duration versus sleep insufficiency as predictors of cardiometabolic health outcomes. Sleep Medicine. 2012;13(10):1261-1270. doi:10.1016/j.sleep.2012.08.005 Gradisar M, Gardner G, Dohnt H. Recent worldwide sleep patterns and problems during adolescence: A review and meta-analysis of age, region, and sleep. Sleep Medicine. 2011;12(2):110-118. doi:10.1016/j.sleep.2010.11.008 Owens J, ADOLESCENT SLEEP WORKING GROUP, COMMITTEE ON ADOLESCENCE, et al. Insufficient Sleep in Adolescents and Young Adults: An Update on Causes and Consequences. Pediatrics. 2014;134(3):e921-e932. doi:10.1542/peds.2014-1696 Lim DC, Najafi A, Afifi L, et al. The need to promote sleep health in public health agendas across the globe. The Lancet Public Health. 2023;8(10):e820-e826. doi:10.1016/S2468-2667(23)00182-2 Kirshenbaum JS, Coury SM, Colich NL, Manber R, Gotlib IH. Objective and subjective sleep health in adolescence: Associations with puberty and affect. Journal of Sleep Research. 2023;32(3):e13805. doi:10.1111/jsr.13805 Yan YW, Lin RM, Su YK, Liu MY. The relationship between adolescent academic stress and sleep quality: A multiple mediation model. soc behav pers. 2018;46(1):63-77. doi:10.2224/sbp.6530 Freeman D, Sheaves B, Waite F, Harvey AG, Harrison PJ. Sleep disturbance and psychiatric disorders. The Lancet Psychiatry. 2020;7(7):628-637. doi:10.1016/S2215-0366(20)30136-X Troynikov O, Watson CG, Nawaz N. Sleep environments and sleep physiology: A review. Journal of Thermal Biology. 2018;78:192-203. doi:10.1016/j.jtherbio.2018.09.012 Knutson KL, Van Cauter E. Associations between Sleep Loss and Increased Risk of Obesity and Diabetes. Annals of the New York Academy of Sciences. 2008;1129(1):287-304. doi:10.1196/annals.1417.033 Prather AA, Leung CW. Association of Insufficient Sleep With Respiratory Infection Among Adults in the United States. JAMA Intern Med. 2016;176(6):850. doi:10.1001/jamainternmed.2016.0787 Sun X, Liu B, Liu S, et al. Sleep disturbance and psychiatric disorders: a bidirectional Mendelian randomisation study. Epidemiol Psychiatr Sci. 2022;31:e26. doi:10.1017/S2045796021000810 Gordon AM, Carrillo B, Barnes CM. Sleep and social relationships in healthy populations: A systematic review. Sleep Medicine Reviews. 2021;57:101428. doi:10.1016/j.smrv.2021.101428 Mah CD, Mah KE, Kezirian EJ, Dement WC. The Effects of Sleep Extension on the Athletic Performance of Collegiate Basketball Players. Sleep. 2011;34(7):943-950. doi:10.5665/SLEEP.1132 Jackson ML, Gunzelmann G, Whitney P, et al. Deconstructing and reconstructing cognitive performance in sleep deprivation. Sleep Medicine Reviews. 2013;17(3):215-225. doi:10.1016/j.smrv.2012.06.007 Hill ED, Kashyap P, Raffanello E, et al. Prediction of mental health risk in adolescents. Nat Med. Published online March 5, 2025. doi:10.1038/s41591-025-03560-7 Grandner MA, Patel NP, Gehrman PR, Perlis ML, Pack AI. Problems associated with short sleep: Bridging the gap between laboratory and epidemiological studies. Sleep Medicine Reviews. 2010;14(4):239-247. doi:10.1016/j.smrv.2009.08.001 He Y, Jones CR, Fujiki N, et al. The Transcriptional Repressor DEC2 Regulates Sleep Length in Mammals. Science. 2009;325(5942):866-870. doi:10.1126/science.1174443 Mayne SL, Mitchell JA, Virudachalam S, Fiks AG, Williamson AA. Neighborhood environments and sleep among children and adolescents: A systematic review. Sleep Medicine Reviews. 2021;57:101465. doi:10.1016/j.smrv.2021.101465 Ohi K, Ochi R, Noda Y, et al. Polygenic risk scores for major psychiatric and neurodevelopmental disorders contribute to sleep disturbance in childhood: Adolescent Brain Cognitive Development (ABCD) Study. Transl Psychiatry. 2021;11(1):187. doi:10.1038/s41398-021-01308-8 Cooper R, Di Biase MA, Bei B, Quach J, Cropley V. Associations of Changes in Sleep and Emotional and Behavioral Problems From Late Childhood to Early Adolescence. JAMA Psychiatry. 2023;80(6):585. doi:10.1001/jamapsychiatry.2023.0379 Zhang L, Sasser J, Doane LD, Peltz J, Oshri A. Latent Profiles of Sleep Patterns in Early Adolescence: Associations With Behavioral Health Risk. Journal of Adolescent Health. 2024;74(1):177-185. doi:10.1016/j.jadohealth.2023.08.021 Sehgal A, Mignot E. Genetics of Sleep and Sleep Disorders. Cell. 2011;146(2):194-207. doi:10.1016/j.cell.2011.07.004 Jha PK, Valekunja UK, Ray S, Nollet M, Reddy AB. Single-cell transcriptomics and cell-specific proteomics reveals molecular signatures of sleep. Commun Biol. 2022;5(1):846. doi:10.1038/s42003-022-03800-3 Krause AJ. The sleep-deprived human brain. Cheng W, Rolls E, Gong W, et al. Sleep duration, brain structure, and psychiatric and cognitive problems in children. Mol Psychiatry. 2021;26(8):3992-4003. doi:10.1038/s41380-020-0663-2 Yang FN, Xie W, Wang Z. Effects of sleep duration on neurocognitive development in early adolescents in the USA: a propensity score matched, longitudinal, observational study. The Lancet Child & Adolescent Health. 2022;6(10):705-712. doi:10.1016/S2352-4642(22)00188-2 Jalbrzikowski M, Hayes RA, Scully KE, et al. Associations between brain structure and sleep patterns across adolescent development. Sleep. 2021;44(10):zsab120. doi:10.1093/sleep/zsab120 Dutil C, Walsh JJ, Featherstone RB, et al. Influence of sleep on developing brain functions and structures in children and adolescents: A systematic review. Sleep Medicine Reviews. 2018;42:184-201. doi:10.1016/j.smrv.2018.08.003 Wang Y, Genon S, Dong D, et al. Covariance patterns between sleep health domains and distributed intrinsic functional connectivity. Nat Commun. 2023;14(1):7133. doi:10.1038/s41467-023-42945-5 Ma Q, Sahakian BJ, Zhang B, Li Z. Neural correlates of device-based sleep characteristics in adolescents. Bresser T, Blanken TF, De Lange SC, et al. Insomnia Subtypes Have Differentiating Deviations in Brain Structural Connectivity. Biological Psychiatry. 2025;97(3):302-312. doi:10.1016/j.biopsych.2024.06.014 Young AL. Uncovering the heterogeneity and temporal complexity of neurodegenerative diseases with Subtype and Stage Inference. NATURE COMMUNICATIONS. Published online 2018. Hirshkowitz M, Whiton K, Albert SM, et al. National Sleep Foundation’s updated sleep duration recommendations: final report. Sleep Health. 2015;1(4):233-243. doi:10.1016/j.sleh.2015.10.004 Paruthi S, Brooks LJ, D’Ambrosio C, et al. Recommended Amount of Sleep for Pediatric Populations: A Consensus Statement of the American Academy of Sleep Medicine. Journal of Clinical Sleep Medicine. 2016;12(06):785-786. doi:10.5664/jcsm.5866 Kwon M, Page SD, Williamson AA, Morgan S, Sawyer AM. Social determinants of health at multiple socio-ecological levels and sleep health in adolescents: A scoping review. Sleep Medicine Reviews. 2024;78:102008. doi:10.1016/j.smrv.2024.102008 Meltzer LJ, Williamson AA, Mindell JA. Pediatric sleep health: It matters, and so does how we define it. Sleep Medicine Reviews. 2021;57:101425. doi:10.1016/j.smrv.2021.101425 Newton AT, Honaker SM, Reid GJ. Risk and protective factors and processes for behavioral sleep problems among preschool and early school-aged children: A systematic review. Sleep Medicine Reviews. 2020;52:101303. doi:10.1016/j.smrv.2020.101303 Tarokh L. Sleep in adolescence: Physiology, cognition and mental health. Neuroscience and Biobehavioral Reviews. Published online 2016. Holm MC, Leonardsen EH, Beck D, et al. Linking brain maturation and puberty during early adolescence using longitudinal brain age prediction in the ABCD cohort. Developmental Cognitive Neuroscience. 2023;60:101220. doi:10.1016/j.dcn.2023.101220 Dashti HS, Jones SE, Wood AR, et al. Genome-wide association study identifies genetic loci for self-reported habitual sleep duration supported by accelerometer-derived estimates. Nat Commun. 2019;10(1):1100. doi:10.1038/s41467-019-08917-4 Sadeh A. Consequences of Sleep Loss or Sleep Disruption in Children. Sleep Medicine Clinics. 2007;2(3):513-520. doi:10.1016/j.jsmc.2007.05.012 Embang JEG, Tan YHV, Ng YX, et al. Role of sleep and neurochemical biomarkers in synaptic plasticity related to neurological and psychiatric disorders: A scoping review. Medic G, Wille M, Hemels ME. Short- and long-term health consequences of sleep disruption. Nature and Science of Sleep. Matricciani L, Blunden S, Rigney G, Williams MT, Olds TS. Children’s Sleep Needs: Is There Sufficient Evidence to Recommend Optimal Sleep for Children? Sleep. 2013;36(4):527-534. doi:10.5665/sleep.2538 Feinberg I. Recommended Sleep Durations for Children and Adolescents: The Dearth of Empirical Evidence. Sleep. 2013;36(4):461-462. doi:10.5665/sleep.2520 Ashbrook LH. Genetics of the human circadian clock and sleep homeostat. Published online 2020. Gomes Da Silva S, Arida RM. Physical activity and brain development. Expert Review of Neurotherapeutics. 2015;15(9):1041-1051. doi:10.1586/14737175.2015.1077115 Valkenborghs SR, Noetel M, Hillman CH, et al. The Impact of Physical Activity on Brain Structure and Function in Youth: A Systematic Review. Pediatrics. 2019;144(4):e20184032. doi:10.1542/peds.2018-4032 Tuan LH, Tsao CY, Lee LJH, Lee LJ. Voluntary exercise ameliorates synaptic pruning deficits in sleep-deprived adolescent mice. Brain, Behavior, and Immunity. 2021;93:96-110. doi:10.1016/j.bbi.2020.12.017 Galland BC, Short MA, Terrill P, et al. Establishing normal values for pediatric nighttime sleep measured by actigraphy: a systematic review and meta-analysis. Sleep. 2018;41(4). doi:10.1093/sleep/zsy017 Olds T, Maher C, Blunden S, Matricciani L. Normative Data on the Sleep Habits of Australian Children and Adolescents. 2010;33(10). Zeringue MM, Saini EK, Fuller-Rowell TE, Hinnant JB, El-Sheikh M. Neighborhood environment and adolescent sleep: The role of family socioeconomic status. Sleep Medicine. 2023;109:40-49. doi:10.1016/j.sleep.2023.06.014 Hunter JC, Hayden KM. The association of sleep with neighborhood physical and social environment. Public Health. 2018;162:126-134. doi:10.1016/j.puhe.2018.05.003 Maret S, Faraguna U, Nelson AB, Cirelli C, Tononi G. Sleep and waking modulate spine turnover in the adolescent mouse cortex. Nat Neurosci. 2011;14(11):1418-1420. doi:10.1038/nn.2934 Huttenlocher PR, Dabholkar AS. Regional differences in synaptogenesis in human cerebral cortex. Journal of Comparative Neurology. 1997;387(2):167-178. doi:10.1002/(SICI)1096-9861(19971020)387:23.0.CO;2-Z Di T, Zhang L, Meng S, et al. The impact of REM sleep loss on human brain connectivity. Transl Psychiatry. 2024;14(1):270. doi:10.1038/s41398-024-02985-x LeGates TA, Fernandez DC, Hattar S. Light as a central modulator of circadian rhythms, sleep and affect. Nat Rev Neurosci. 2014;15(7):443-454. doi:10.1038/nrn3743 Li Y, Gong W, Sahakian BJ, et al. Divergent biological pathways linking short and long sleep durations to mental and physical health. Nat Mental Health. 2025;3(4):429-443. doi:10.1038/s44220-025-00395-6 Keshavan MS, Giedd J, Lau JYF, Lewis DA, Paus T. Changes in the adolescent brain and the pathophysiology of psychotic disorders. Lancet Psychiatry. 2014;1(7):549-558. doi:10.1016/S2215-0366(14)00081-9 Paus T, Keshavan M, Giedd JN. Why do many psychiatric disorders emerge during adolescence? Nat Rev Neurosci. 2008;9(12):947-957. doi:10.1038/nrn2513 Solmi M, Radua J, Olivola M, et al. Age at onset of mental disorders worldwide: large-scale meta-analysis of 192 epidemiological studies. Mol Psychiatry. 2022;27(1):281-295. doi:10.1038/s41380-021-01161-7 Garcia AD, Buffalo EA. Anatomy and Function of the Primate Entorhinal Cortex. Annu Rev Vis Sci. 2020;6:411-432. doi:10.1146/annurev-vision-030320-041115 Takehara-Nishiuchi K, Maal-Bared G, Morrissey MD. Increased Entorhinal–Prefrontal Theta Synchronization Parallels Decreased Entorhinal–Hippocampal Theta Synchronization during Learning and Consolidation of Associative Memory. Front Behav Neurosci. 2012;5:90. doi:10.3389/fnbeh.2011.00090 Knopman DS, Lundt ES, Therneau TM, et al. Entorhinal cortex tau, amyloid-β, cortical thickness and memory performance in non-demented subjects. Brain. 2019;142(4):1148-1160. doi:10.1093/brain/awz025 Schmaal L, Hibar DP, Sämann PG, et al. Cortical abnormalities in adults and adolescents with major depression based on brain scans from 20 cohorts worldwide in the ENIGMA Major Depressive Disorder Working Group. Mol Psychiatry. 2017;22(6):900-909. doi:10.1038/mp.2016.60 Shaw P, Eckstrand K, Sharp W, et al. Attention-deficit/hyperactivity disorder is characterized by a delay in cortical maturation. Proceedings of the National Academy of Sciences. 2007;104(49):19649-19654. doi:10.1073/pnas.0707741104 Luby JL, Belden AC, Jackson JJ, et al. Early Childhood Depression and Alterations in the Trajectory of Gray Matter Maturation in Middle Childhood and Early Adolescence. JAMA Psychiatry. 2016;73(1):31-38. doi:10.1001/jamapsychiatry.2015.2356 Meyer N, Harvey AG, Lockley SW, Dijk DJ. Circadian rhythms and disorders of the timing of sleep. The Lancet. 2022;400(10357):1061-1078. doi:10.1016/S0140-6736(22)00877-7 Eylon G, Tikotzky L, Dinstein I. Performance evaluation of Fitbit Charge 3 and actigraphy vs. polysomnography: Sensitivity, specificity, and reliability across participants and nights. Sleep Health. 2023;9(4):407-416. doi:10.1016/j.sleh.2023.04.001 Aili K, Åström-Paulsson S, Stoetzer U, Svartengren M, Hillert L. Reliability of Actigraphy and Subjective Sleep Measurements in Adults: The Design of Sleep Assessments. Journal of Clinical Sleep Medicine. 2017;13(01):39-47. doi:10.5664/jcsm.6384 Lin J, Sun W, Song Y, et al. Cohort Profile: The Shanghai Sleep Birth Cohort Study. Paediatric Perinatal Epid. 2021;35(2):257-268. doi:10.1111/ppe.12738 Dale AM, Fischl B, Sereno MI. Cortical Surface-Based Analysis: I. Segmentation and Surface Reconstruction. NeuroImage. 1999;9(2):179-194. doi:10.1006/nimg.1998.0395 Casey BJ, Cannonier T, Conley MI, et al. The Adolescent Brain Cognitive Development (ABCD) study: Imaging acquisition across 21 sites. Developmental Cognitive Neuroscience. 2018;32:43-54. doi:10.1016/j.dcn.2018.03.001 Desikan RS, Segonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Published online 2006. Pomponio R, Erus G, Habes M, et al. Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan. NeuroImage. 2020;208:116450. doi:10.1016/j.neuroimage.2019.116450 Bruni O, Ottaviano S, Guidetti V, et al. The Sleep Disturbance Scale for Children (SDSC) Construct ion and validation of an instrument to evaluate sleep disturbances in childhood and adolescence. Journal of Sleep Research. 1996;5(4):251-261. doi:10.1111/j.1365-2869.1996.00251.x Owens JA, Spirito A, McGuinn M. The Children’s Sleep Habits Questionnaire (CSHQ): Psychometric Properties of A Survey Instrument for School-Aged Children. Sleep. 2000;23(8):1-9. doi:10.1093/sleep/23.8.1d Zavada A, Gordijn MCM, Beersma DGM, Daan S, Roenneberg T. Comparison of the Munich Chronotype Questionnaire with the Horne‐Östberg’s Morningness‐Eveningness score. Chronobiology International. 2005;22(2):267-278. doi:10.1081/CBI-200053536 Cardenas-Iniguez C, Schachner JN, Ip KI, et al. Building towards an adolescent neural urbanome: Expanding environmental measures using linked external data (LED) in the ABCD study. Developmental Cognitive Neuroscience. 2024;65:101338. doi:10.1016/j.dcn.2023.101338 Achenbach TM, Ruffle TM. The Child Behavior Checklist and Related Forms for Assessing Behavioral/Emotional Problems and Competencies. Pediatrics In Review. 2000;21(8):265-271. doi:10.1542/pir.21.8.265 Goodman R. The Strengths and Difficulties Questionnaire: A Research Note. Child Psychology Psychiatry. 1997;38(5):581-586. doi:10.1111/j.1469-7610.1997.tb01545.x Goodman A, Lamping DL, Ploubidis GB. When to Use Broader Internalising and Externalising Subscales Instead of the Hypothesised Five Subscales on the Strengths and Difficulties Questionnaire (SDQ): Data from British Parents, Teachers and Children. J Abnorm Child Psychol. 2010;38(8):1179-1191. doi:10.1007/s10802-010-9434-x Sun H, Mehta S, Khaitova M, et al. Brain age prediction and deviations from normative trajectories in the neonatal connectome. Nat Commun. 2024;15(1):10251. doi:10.1038/s41467-024-54657-5 Shen X, Finn ES, Scheinost D, et al. Using connectome-based predictive modeling to predict individual behavior from brain connectivity. Nat Protoc. 2017;12(3):506-518. doi:10.1038/nprot.2016.178 Hansen JY, Shafiei G, Markello RD, et al. Mapping neurotransmitter systems to the structural and functional organization of the human neocortex. Nat Neurosci. 2022;25(11):1569-1581. doi:10.1038/s41593-022-01186-3 Dukart J, Holiga S, Rullmann M, et al. JuSpace: A tool for spatial correlation analyses of magnetic resonance imaging data with nuclear imaging derived neurotransmitter maps. Choi SW, Mak TSH, O’Reilly PF. Tutorial: a guide to performing polygenic risk score analyses. Nat Protoc. 2020;15(9):2759-2772. doi:10.1038/s41596-020-0353-1 Gelman A, Hwang J, Vehtari A. Understanding predictive information criteria for Bayesian models. Stat Comput. Published online 2014. Additional Declarations There is NO Competing Interest. Supplementary Files supplement1.docx Supplement 1 supplement2.docx Supplement 2 Cite Share Download PDF Status: Published Journal Publication published 07 Mar, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6851158","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":475329051,"identity":"8ccf1eb4-f633-4979-af0d-7e64a19f175e","order_by":0,"name":"Dan Wu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIiWNgGAWjYDCCAzCSvYHBAMhibCBeC88BkrVIJIBZhLXwHT97TJqH4U40v+TzB8U8DDayGw4wP3uAT4vkmbw0oJZnuTNnJyQY8zCkGW84wGZugE+LwYEcs9s8DIdzN9xOOADUcjhxwwEeNgm8Ws6/gWjZf/NgA1DLfyK03IDZIsHMANRygLAWyRtvzH/OMXiWO+NMGoPhHINk45mH2czwauE7n2Ns8KbiTm5/+/FnQIadbN/x5md4tUCdBybZDMAMZsLq4YD5AQmKR8EoGAWjYAQBAByRTIiSUofsAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-9303-5821","institution":"Zhejiang University","correspondingAuthor":true,"prefix":"","firstName":"Dan","middleName":"","lastName":"Wu","suffix":""},{"id":475329052,"identity":"8e40acca-6286-40ff-82ff-00e541f06d41","order_by":1,"name":"Yiwei Chen","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Yiwei","middleName":"","lastName":"Chen","suffix":""},{"id":475329053,"identity":"c1ec96b4-5349-41b8-bb52-6ab5e42ebdad","order_by":2,"name":"Mingyang Li","email":"","orcid":"","institution":"Zhejiang university","correspondingAuthor":false,"prefix":"","firstName":"Mingyang","middleName":"","lastName":"Li","suffix":""},{"id":475329054,"identity":"4e034a3c-3ff2-40d9-a026-b09b5e32e069","order_by":3,"name":"Zhiyong Zhao","email":"","orcid":"https://orcid.org/0000-0002-1432-0430","institution":"Department of Biomedical Engineering, Children's Hospital, Zhejiang University School of Medicine, National Clinical Research Center for Child Health, Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Zhiyong","middleName":"","lastName":"Zhao","suffix":""},{"id":475329055,"identity":"8c9988ea-d7b9-483e-a591-c79d85195a8e","order_by":4,"name":"Xinyi Xu","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Xinyi","middleName":"","lastName":"Xu","suffix":""},{"id":475329056,"identity":"7c7b72a7-8241-4f95-8f2d-728d25704334","order_by":5,"name":"Yongquan Huang","email":"","orcid":"","institution":"Zhejiang university","correspondingAuthor":false,"prefix":"","firstName":"Yongquan","middleName":"","lastName":"Huang","suffix":""},{"id":475329057,"identity":"5737211d-2e64-4388-b145-4c4acbd2cdb8","order_by":6,"name":"Ruike Chen","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Ruike","middleName":"","lastName":"Chen","suffix":""},{"id":475329058,"identity":"1065e8a8-3fa8-420a-84d2-83ed8bc28576","order_by":7,"name":"Ruoke Zhao","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Ruoke","middleName":"","lastName":"Zhao","suffix":""},{"id":475329059,"identity":"78748ed2-1b90-4585-9dfb-7e378946810b","order_by":8,"name":"Guanghai Wang","email":"","orcid":"","institution":"Shanghai Jiao Tong University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Guanghai","middleName":"","lastName":"Wang","suffix":""},{"id":475329060,"identity":"5b6e22c1-b7c0-4c6a-80f7-d9ea21d8242b","order_by":9,"name":"Fan Jiang","email":"","orcid":"https://orcid.org/0000-0003-0634-101X","institution":"Shanghai Children’s Medical Center affiliated Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Fan","middleName":"","lastName":"Jiang","suffix":""}],"badges":[],"createdAt":"2025-06-09 06:20:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6851158/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6851158/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-026-70135-6","type":"published","date":"2026-03-07T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":85372095,"identity":"9cdbe0ab-6067-4830-9cfa-21b92b3a049a","added_by":"auto","created_at":"2025-06-25 07:41:03","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":506395,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverview of the present study. (a)\u003c/strong\u003e, The discovery samples of the ABCD second-year follow-up study with Fitbit sleep measurements. The group difference map showed reduced cortical thickness for insufficient sleep adolescents. \u003cstrong\u003e(b)\u003c/strong\u003e, Individuals with insufficient sleep were classified according to the sequence of cortical thickness alteration in different brain regions. \u003cstrong\u003e(c)\u003c/strong\u003e, The longitudinal change of cortical thickness in the three insufficient sleep subtypes mirrored that of the progression subtyping results. \u003cstrong\u003e(d-f)\u003c/strong\u003e, Three subtypes had distinctive social-ecological factors in terms of sleep factor presentations (d), mental conditions (e), and neighborhood environment factor presentations (f). \u003cstrong\u003e(g)\u003c/strong\u003e, The three subtypes were replicated in an independent Shanghai dataset. \u003cstrong\u003e(h)\u003c/strong\u003e Summary of this study. The three subtypes were characterized by distinct socio-ecological factors, namely, the natural short sleepers, the environment-related, and the comorbidity-related subgroups\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/e3a417e7e7b78575f2c5e500.png"},{"id":85372388,"identity":"7e7f7784-cd22-49d6-8ab8-eab3f9495434","added_by":"auto","created_at":"2025-06-25 07:49:03","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":581646,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBrain morphology of the three subtypes of insufficient sleep. (a)\u003c/strong\u003e, ROI-based group differences between insufficient and sufficient sleep adolescents. L = left; R = right. \u003cstrong\u003e(b)\u003c/strong\u003e, Cortical thickness (CT) developmental ‘trajectories’ in the three subgroups. The positional variance diagrams employ varying colors to depict the probability of each brain region attaining a specific z-score. Different colors correspond with varying degrees of reduced CT: mildly reduced regions are shown in red (z = 1), moderately reduced regions in magenta (z = 1.5), and severely reduced regions in blue (z = 2). Colors indicate the proportion of the posterior distribution where events occur in a particular position in the sequence. \u003cstrong\u003e(c)\u003c/strong\u003e, Probability-based distribution of subjects in the ‘post-effect’ phase of the three subtypes. \u003cstrong\u003e(d)\u003c/strong\u003e, Z-score maps of the difference in cortical thickness (CT) between insufficient and sufficient sleep groups at each stage of the three subtypes. Stage I (stage = 1), stage II (2 ≤ stage ≤ 3), stage III (4 ≤ stage ≤ 7), stage IV (8 ≤ stage ≤ 15), stage V (16 ≤ stage).\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/8147128e70efe926a8c81776.png"},{"id":85372098,"identity":"c6a99a90-4144-4457-9725-b2a08717e8e0","added_by":"auto","created_at":"2025-06-25 07:41:03","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":715342,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLongitudinal change of cortical thickness in the three sleep insufficiency subtypes. (a)\u003c/strong\u003e, The insufficient sleep group showed significantly larger within-group variability in the longitudinal change of cortical thickness (ΔCT, measured as the difference between baseline and follow-up visit), compared to the sufficient sleep group. \u003cstrong\u003e(b)\u003c/strong\u003e, The postcentral subtype showed similar within-group variability compared with the sufficient sleep group. \u003cstrong\u003e(c)\u003c/strong\u003e, Difference in ΔCT between each subtype and the controls. \u003cstrong\u003e(d)\u003c/strong\u003e, Subtype brain morphology pseudo-progression pattern found with the SuStaIn model, which mirrored the actual cortical thickness progression. Pseudo-progression brain maps were obtained by subtracting subtypes’ stage I z-score brain maps from stage III z-score brain maps.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/db6e9cc2191fd3e4d613a6e3.png"},{"id":85372100,"identity":"81f58c3f-e242-4c8d-907b-a5a6ed659962","added_by":"auto","created_at":"2025-06-25 07:41:03","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":281875,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSocio-ecological and clinical characterization of three insufficient sleep subtypes. (a)\u003c/strong\u003e, Pediatric sleep health, mental or physical conditions, and family and environment factors are three layers of socio-ecological factors for pediatric sleep\u003csup\u003e35\u003c/sup\u003e. \u003cstrong\u003e(b)\u003c/strong\u003e, According to the three layers of socio-ecological factors, the postcentral subtype had similar presentations as control; the pericalcarine subtype was under adverse environmental conditions (brighter night light, louder noise at night, less neighborhood income, lower neighborhood education, and more serious neighborhood disadvantage), worse physical activities, and had sleep problems (insomnia, going to bed reluctancy, and weekly sleep loss) compared with control; while the entorhinal subtype was mostly affected in physical and mental conditions, with lower physical activities, higher correlations with mental disorders, and worse pediatric sleep domain factors (insomnia, sleep latency, going to bed reluctance, falling asleep difficulty, falling asleep anxiety, and weekly sleep loss). Values for each bar in the graph are summarized in the Supplementary Table S2.(\u003cstrong\u003ec\u003c/strong\u003e), Heat map comparisons showing the group differences (Cohen’d or Rho) between insufficient sleep subtypes and control in terms of sleep socio-ecological domains (pediatric sleep, physical and mental factors, and neighborhood factors). For externalizing, internalizing, and overall problem scores \u003cstrong\u003e(d)\u003c/strong\u003e, A graphical summary of socio-ecological differences of different subtypes, with the postcentral subtype showing natural short sleeper-like presentations, the pericalcarine subtype showing environmentally related presentations, and the entorhinal subtype showing psychiatric comorbid related presentations. Abbreviations: DIMS — disorders of initiating and maintaining sleep; SBD — sleep breathing disorders; DA — disorder of arousal nightmares; SWTD — sleep wake transition disorders; DOES — disorder of excessive somnolence; SHY — sleep hyperhidrosis.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/0ca5619e4a29832ff128f0b7.png"},{"id":85372392,"identity":"ce1e26dc-4483-4760-a400-ec4e129c917a","added_by":"auto","created_at":"2025-06-25 07:49:03","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":233578,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSocio-ecological signatures of the three subtypes in terms of brain age, environmental effect, cognition-behavior, and neurotransmitters. (a)\u003c/strong\u003e, The brain age of the three subtypes and the control group was compared. The postcentral subtype (AKA. short sleepers) had significantly higher brain age, and the entorhinal (AKA. Comorbid related) subtype had lower brain age compared with the control. \u003cstrong\u003e(b)\u003c/strong\u003e, The short sleep PRS score of the three subtypes and the control group was compared. The short sleepers had significantly larger PRS score, indicating a genetic predisposition for short sleep duration. \u003cstrong\u003e(c-e), \u003c/strong\u003eFitbit measured sleep duration significantly mediated the relationship between nightlight and pericalcarine (visual cortex) cortical thickness of three subtypes. Only the pericalcarine subtype was found with sleep duration mediating the association between nightlight and visual cortex CT \u003cstrong\u003e(d)\u003c/strong\u003e. \u003cstrong\u003e(f-h)\u003c/strong\u003e, Associations between SuStaIn stage and residualized CBCL score, i.e., externalizing problems \u003cstrong\u003e(f)\u003c/strong\u003e, internalizing problems \u003cstrong\u003e(g),\u003c/strong\u003e and overall problems \u003cstrong\u003e(h)\u003c/strong\u003e. SuStaIn stage of the entorhinal subtype showed significant correlations with the severity of externalizing, internalizing, and overall problems, which was not found in other subtypes. The shaded region denotes the 95% confidence interval enveloping the regression line.\u003cstrong\u003e (i-k)\u003c/strong\u003e Spatial correlations between neurotransmitter and cortical thickness profiles of the three subtypes. * P\u003csub\u003epermu\u003c/sub\u003e\u0026lt;0.001. The dashed-dotted line stands for Spearman Rho \u0026gt; 0.5.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/f197f2d64c48b15e7f801ef2.png"},{"id":85373610,"identity":"62abe73a-fcbd-4864-842c-5cf943180203","added_by":"auto","created_at":"2025-06-25 08:05:03","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":312410,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eValidation of the insufficient sleep subtypes with the SSBC dataset. (a-c)\u003c/strong\u003e, A replication of insufficient sleep-induced change of cortical thickness in the three subtypes compared with sufficient sleep group in the SSBC dataset. \u003cstrong\u003e(d)\u003c/strong\u003e, Pediatric sleep presentation of different subtypes. The pericalcarine and the entorhinal subtypes had worse pediatric sleep presentations than the controls, while the postcentral subtype had similar presentations. Values for each bar in the graph are summarized in the Supplementary Table S3. \u003cstrong\u003e(e-g)\u003c/strong\u003e, Associations between SuStaIn stage and the residualized CBCL score including externalizing problems \u003cstrong\u003e(e)\u003c/strong\u003e, internalizing problems \u003cstrong\u003e(f),\u003c/strong\u003e and overall problems \u003cstrong\u003e(g)\u003c/strong\u003e. Only the SuStaIn stage in the entorhinal subtype (green line) showed significant correlations with the severity of externalizing and overall problems. The shaded region denotes the 95% confidence interval enveloping the regression line.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/1a6a201e01078b9ce2f3ee44.png"},{"id":107398141,"identity":"37b8f7c9-84eb-4ab0-a45d-de641bf5525f","added_by":"auto","created_at":"2026-04-21 07:05:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3107339,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/67890a62-29ba-4830-859a-a93da2b0ba55.pdf"},{"id":85372092,"identity":"a1dcb6d4-bb5e-4b15-8bf0-3239f9d1d6ac","added_by":"auto","created_at":"2025-06-25 07:41:03","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":906954,"visible":true,"origin":"","legend":"Supplement 1","description":"","filename":"supplement1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/5c17c4fae821477a9157b3b3.docx"},{"id":85372387,"identity":"9a442e5f-27b1-4e0f-bac0-6ab7e70799e5","added_by":"auto","created_at":"2025-06-25 07:49:03","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":17180,"visible":true,"origin":"","legend":"Supplement 2","description":"","filename":"supplement2.docx","url":"https://assets-eu.researchsquare.com/files/rs-6851158/v1/cdcaac428034e4832a40bce6.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Heterogeneous Neurophenotypes of Adolescent Sleep Insufficiency Stratify Natural Short Sleepers from Comorbidity or Environment Driven Insufficiency","fulltext":[{"header":"Introduction","content":"\u003cp\u003eInsufficient sleep is prevalent in modern society, affecting 1/3 of the population\u003csup\u003e1\u003c/sup\u003e. Particularly, sleep problems in adolescents have become a major concern due to changes in the social environment, which are complicated by rapid neurobiological development\u003csup\u003e2\u003c/sup\u003e. According to previous studies, adolescent sleep of less than 8 hours is defined as insufficient sleep, affecting 16% of sixth graders\u003csup\u003e3\u003c/sup\u003e. A good night's sleep is hard to have, as it is affected by various factors\u003csup\u003e4\u003c/sup\u003e, including physiological changes of puberty\u003csup\u003e5\u003c/sup\u003e, academic stress load\u003csup\u003e6\u003c/sup\u003e, mental health\u003csup\u003e7\u003c/sup\u003e, and sleeping environments\u003csup\u003e8\u003c/sup\u003e. Insufficient sleep also causes major social and health consequences, including both physical\u003csup\u003e9,10\u003c/sup\u003e and mental fatigue\u003csup\u003e11\u003c/sup\u003e, negatively influencing social life\u003csup\u003e12\u003c/sup\u003e and school performance\u003csup\u003e13,14\u003c/sup\u003e. A recent study reported that sleep disturbances emerged as the most influential predictor of mental health risk\u003csup\u003e15\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe multifaceted and heterogeneous determinants and outcomes of insufficient sleep hinder our understanding of sleep in children and adolescents\u003csup\u003e16\u003c/sup\u003e. Previous studies used the social-ecological framework to understand the etiology of sleep problems. Under such a framework, causes of insufficient sleep were divided into three layers, namely, sleep-related factors, personal conditions (mental or physical discomfort), and environmental factors. In practice, some children may be biologically disposed to short sleep duration without compromised development\u003csup\u003e17\u003c/sup\u003e, while others sleep less because of external disturbance\u003csup\u003e18\u003c/sup\u003e, psychiatric comorbidity\u003csup\u003e19\u003c/sup\u003e, or other conditions, that require entirely different interventions. To settle such a problem, a reliable, unbiased, and informative biomarker becomes the key for stratification.\u003c/p\u003e \u003cp\u003eTo capture the different sleep profiles, previous studies focused on various objective and subjective sleep measures using latent profile analysis \u003csup\u003e20,21\u003c/sup\u003e. Other studies of sleep genetics\u003csup\u003e22\u003c/sup\u003e and proteomics\u003csup\u003e23\u003c/sup\u003e found genes and proteins for sleep regulation. On the other hand, as insufficient sleep has major impacts on brain development\u003csup\u003e24\u003c/sup\u003e, an in-depth understanding of neurobiology needs to be further addressed. In recent years, MRI-based neuroimaging approaches have been used frequently to unravel the underlying neurological mechanisms of sleep problems. Cross-sectional studies found that shorter sleep time is correlated with lower cortical brain volume for adolescents\u003csup\u003e25,26\u003c/sup\u003e. A longitudinal study confirmed that shorter objective sleep time is correlated with a thinner cortex\u003csup\u003e27\u003c/sup\u003e. A review summarized that insufficient sleep strongly correlates with neuroanatomical structures and functions of adolescent brain development, but due to the heterogeneity of insufficient sleep, a precise conclusion cannot be drawn\u003csup\u003e28\u003c/sup\u003e. Other studies also found significant correlations between subjective\u003csup\u003e29\u003c/sup\u003e or objective\u003csup\u003e30\u003c/sup\u003e sleep measures and distinctive brain phenotypes, further confirming the presence of insufficient sleep brain subtypes. Therefore, although at the group level, the current findings pointed to cortical thinning and reduced cortical volumes in the adolescent population with insufficient sleep, the substantial inter-personal differences among individuals\u003csup\u003e16\u003c/sup\u003e demand better stratification.\u003c/p\u003e \u003cp\u003eThe rich anatomical and functional information and the unbiased nature of neuroimaging make it an ideal neurobiological marker for the stratification of sleep disorders. For instance, previous studies found different insomnia subtypes in adults with distinctive cortical topography and structural connectivity\u003csup\u003e31\u003c/sup\u003e. It is not yet known whether individuals with insufficient sleep (i.e., short sleep duration) exhibit subgroups with distinct brain patterns, especially in the adolescent population, and whether the imaging phenotypes possess explicit social-ecological characteristics that can help to inform effective interventional strategies.\u003c/p\u003e \u003cp\u003eIn this study, we aim to decipher the social-ecological heterogeneity of insufficient sleep in adolescents based on brain MRI, including the biological, mental, and environmental factors. We adopted the Subtype and Stage Inference (SuStaIn)\u003csup\u003e32\u003c/sup\u003e method to identify insufficient sleep subtypes based on cortical thickness (CT) using the Adolescent Brain Cognitive Development (ABCD) study with MRI and Fitbit-based sleep measurements. An optimal number of three subtypes was found using the Bayesian model selection and was validated using the longitudinal data from the ABCD dataset and a house-collected independent dataset. To understand the socio-ecological determinants underlying these subtypes, we comprehensively examined pediatric sleep health, physical or mental conditions, and family and neighborhood factors among these subtypes. Lastly, we compared the brain ages, polygenic score (PGS), associations with psychiatric disorder, and neurotransmitter profiles among these subtypes. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e provides an overview of the present study.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003eBrain morphology-based subtyping of insufficient sleep\u003c/h2\u003e\n \u003cp\u003eWe used objective sleep duration measured by Fitbit to characterize insufficient sleep. According to guidelines\u003csup\u003e33,34\u003c/sup\u003e, adolescents with objectively measured sleep of less than 8 hours were defined as insufficient sleepers. We also compared other objective sleep measurements between the insufficient and sufficient sleep groups. The insufficient subtype had significantly higher sleep efficiency (P\u003csub\u003eBonferroni\u003c/sub\u003e \u0026lt; 1x10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e), fewer average wake counts per night (P\u003csub\u003eBonferroni\u003c/sub\u003e \u0026lt; 1x10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e), and similar sleep latency (P\u003csub\u003eBonferroni\u003c/sub\u003e = 0.71).\u003c/p\u003e\n \u003cp\u003eWe first identified the brain patterns most susceptible to sleep deprivation by comparing the insufficient and sufficient sleep groups. Using the Desikan-Killiany (DK) Atlas as a template, we identified the region of interest (ROI) with significantly altered cortical thickness (CT) across the whole-brain after controlling for random factors of age, sex, ethnicity, BMI and site, using subjects from the ABCD second-year follow-up (\u003cem\u003eN\u003c/em\u003e\u003csub\u003e\u003cem\u003einsufficient sleep\u003c/em\u003e\u003c/sub\u003e = 853, \u003cem\u003eN\u003c/em\u003e\u003csub\u003e\u003cem\u003esufficient sleep\u003c/em\u003e\u003c/sub\u003e = 2,413). Figure \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea demonstrated that sleep had a profound effect on the developing adolescent brain. 29 of 68 brain regions showed reduced CT in the insufficient sleep group compared to the controls, with the left lateral orbitofrontal cortex showing the most reduced CT with Cohen\u0026rsquo;s d of 0.200 (P\u003csub\u003eBonferroni\u003c/sub\u003e = 7.95x10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e).\u003c/p\u003e\n \u003cp\u003eWe then used the SuStaIn model to identify distinct longitudinal trajectories of CT reduction with three z-score cutoffs: 1, 1.5, and 2. A three-cluster model stood out with minimum cross-validation information criterion (Supplementary Fig. \u003cspan class=\"InternalRef\"\u003eS2\u003c/span\u003ea) and maximal log-likelihood (Supplementary Fig. \u003cspan class=\"InternalRef\"\u003eS2\u003c/span\u003eb) under ten-fold validation. This resulted in three subtypes with distinctive topography, i.e., subtype 1 showed CT reduction across the whole whole-brain (n\u0026thinsp;=\u0026thinsp;197, 23.09% of the total), subtype 2 showed a lower CT in the occipital lobe (n\u0026thinsp;=\u0026thinsp;396, 46.42% of the total), and subtype 3 showed a lower CT in the entorhinal cortex (n\u0026thinsp;=\u0026thinsp;260, 30.48% of the total). To test the stability of the results, other z-score cutoffs (1, 2, and 3; 2, 3, and 4) for SuStaIn were also tested (Supplementary Fig. S4).\u003c/p\u003e\n \u003cp\u003eSleep deficit-induced brain changes also showed different progressive trajectories among the subtypes. Subtype 1 exhibited initial change in the postcentral cortex, which then progressed to the supramarginal precuneus and inferior parietal area, namely, the postcentral-originated subtype. Subtype 2 showed initial effects in the pericalcarine area, which then progressed to the lingual area, namely, the pericalcarine subtype. Subtype 3 first demonstrated alteration in the entorhinal cortex, then progressed to the inferior temporal and precentral cortex, namely, the entorhinal subtype.\u003c/p\u003e\n \u003cp\u003eThe SuStaIn model assigned each insufficient sleep adolescent with the most likely subtype and stage label. The subjects assigned to \u0026lsquo;stage 0\u0026rsquo; of the progression stage were noted as \u0026lsquo;pre-effect\u0026rsquo;, while others were noted as \u0026lsquo;post-effect\u0026rsquo;. The probability-based distribution of post-effect individuals in the three subtypes was depicted in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ec. Figure \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed illustrated the different stages in the three subtypes of sleep deficit, represented as the mean z-value images, i.e., reduction of CT in the \u0026lsquo;post-effect\u0026rsquo; phase compared to the controls.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eLongitudinal change of brain morphology in the three subtypes\u003c/h3\u003e\n\u003cp\u003eTo measure the change of brain CT and compare it with pseudo staging from SuStaIn, we obtained \u0026Delta;CT by subtracting the CT values between two rounds of longitudinal MRI scans while regressing the age gap between scans and other covariates. We measured the variability of \u0026Delta;CT by log-transformed Euclidean distances within groups, and found the insufficient sleep group had higher within-group variability under multiple atlases (DK and Destrieux atlas, Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea), supporting the higher heterogeneity and the necessity for subtyping. After subtyping, we found the within-group variability of \u0026Delta;CT for the postcentral-originated subtype no longer showed significant difference compared with the sufficient sleep group, indicating the similarity of this subgroup to the controls (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb).\u003c/p\u003e\n\u003cp\u003eFig. 3c demonstrated that the postcentral subtype showed much greater \u0026Delta;CT in the parietal lobe, while the pericalcarine subtype showed greater \u0026Delta;CT in the occipital lobe, and the entorhinal subtype showed greater \u0026Delta;CT in the temporal cortex, which were in accordance with the \u0026ldquo;pseudo progression\u0026rdquo; obtained from SuStaIn model (Fig. 3d).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eCharacterization of subtypes by social-ecological profiles\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePrevious studies used socio-ecological and systems frameworks to understand the determinants of sleep patterns and problems in children and adolescents\u003csup\u003e35\u0026ndash;37\u003c/sup\u003e. To take the layers of socio-ecological factors (Fig. 4a) into full account, we compared the insufficient sleep subtypes with the control group in terms of sleep factors, physical or mental conditions, and neighborhood environment (Fig. 4b), while controlling for confounders including age, sex, BMI, site, race, ethnicity, socioeconomic status. Interestingly, no significant differences were found between the postcentral subtype and the control for any of the socio-ecological factors. Therefore, we interpreted this subtype as natural short sleepers who although sleep relatively less, do not show compromised development, and thus may not require any intervention. We further revealed this group exhibited advanced brain age that may explain the lower CT and increased \u0026Delta;CT in this group. Comparing the polygenetic risk score (PRS) based on GWAS data of short sleep duration, this group also showed a short sleep duration genetic predisposition.\u003c/p\u003e\n\u003cp\u003eThe pericalcarine subtype was found under adverse environmental conditions (brighter night light, louder noise at night, less neighborhood income, lower neighborhood education, and more serious neighborhood disadvantage), and worse physical activities, leading to sleep problems (initiating and maintaining sleep and weekly sleep loss). Therefore, we labeled this subtype as environment-induced sleep deficits. We further performed mediation analysis to investigate the relationship between night light, sleep duration, and pericalcarine in this group.\u003c/p\u003e\n\u003cp\u003eThe entorhinal subtype was mostly affected in pediatric sleep domains, including initiating and maintaining sleep and weekly sleep loss and weekly sleep loss. By examing the linear correlations between SuStaIn stages and residualized CBCL score, we found higher staging in the entorhinal subtype were associated higher externalizing problems (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.394, \u003cem\u003eP\u003c/em\u003e\u003csub\u003eBonferroni\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.008, Bonferroni corrected; Fig. 5f), higher internalizing problems (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.319, \u003cem\u003eP\u003c/em\u003e\u003csub\u003eBonferroni\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.032, Bonferroni corrected; Fig. 5g), and overall problems (r\u0026thinsp;=\u0026thinsp;0.30, \u003cem\u003eP\u003c/em\u003e\u003csub\u003eBonferroni\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.045, Bonferroni corrected; Fig. 5h). Given that this subtype showed higher degree of psychiatric comorbidity compared to the other subtypes, we characterized this group as comorbidity-related. The overall socio-ecological profiles of the subtypes were summarized in Fig. 4d. \u0026nbsp;\u003c/p\u003e\n\u003ch3\u003eBrain age differences among subtypes\u003c/h3\u003e\n\u003cp\u003eDuring adolescence, the need for sleep duration strongly declines with brain maturation\u003csup\u003e38\u003c/sup\u003e. As brain age derived from neuroimaging has been widely used as a biomarker for brain aging and maturation\u003csup\u003e39\u003c/sup\u003e, we tested whether the brain age differed among the subtypes, although the chronological age showed no significant difference (Fig. S7). Compared with the control, the postcentral subtype (aka. short sleepers) showed higher (P\u003csub\u003eBonferroni\u003c/sub\u003e \u0026lt; 0.001, Bonferroni corrected; Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ea) and the entorhinal (aka. comorbid related) showed lower (P\u003csub\u003eBonferroni\u003c/sub\u003e = 0.022, Bonferroni corrected; Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ea) brain age. This result indicated more advanced development in the postcentral subtype (the short-sleepers), which well explained the relatively low need for long sleep duration. The lower brain age in the entorhinal subtype suggested potentially delayed development, potentially due to the psychiatric comorbidity.\u003c/p\u003e\n\u003ch3\u003ePolygenetic risk score of short sleep among subtypes\u003c/h3\u003e\n\u003cp\u003ePrevious studies found a significant genetic basis for short sleepers\u003csup\u003e17,40\u003c/sup\u003e. Hassan et.al\u003csup\u003e40\u003c/sup\u003e found 27 significant loci correlated with short sleepers in adults (subjective sleep\u0026thinsp;\u0026lt;\u0026thinsp;7 hours), and were further validated with accelerometer-measured objective sleep. We calculated the short sleep duration polygenetic risk score (PRS, details see Method). The postcentral subtype had a significantly higher PRS value compared with the control (P\u0026thinsp;=\u0026thinsp;0.0042; Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eb), which indicated a genetic predisposition for short sleep duration. The other insufficient sleep subtypes showed no significant differences in PRS compared with the control.\u003c/p\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003eMediating effects of sleep on the environment and the brain\u003c/h2\u003e\n \u003cp\u003eTo further understand the potential environmental effect on the pericalcarine subtype, we used mediation analysis to determine whether the associations between night light pollution and CT changes in the pericalcarine cortex could be explained by sleep duration. In the pericalcarine subtype, sleep duration mediated the association between nightlight and pericalcarine CT (path C: -0.061, 95% CI [-0.112, -0.010], p\u0026thinsp;=\u0026thinsp;0.022; path a*b: -0.021, 95% CI [-0.037, -0.009], p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ed). The mediation effects were not significant for the postcentral (path C: -0.056, 95% CI [-0.123, 0.010], p\u0026thinsp;=\u0026thinsp;0.094; path a*b: -0.002, 95% CI [-0.008, 0.003], p\u0026thinsp;=\u0026thinsp;0.486; Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ec) or entorhinal (path C: -0.027, 95% CI [-0.091, 0.037], p\u0026thinsp;=\u0026thinsp;0.406; path a*b: 0.004, 95% CI [-0.002, 0.012], p\u0026thinsp;=\u0026thinsp;0.166; Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ee) subtypes.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eNeurotransmitter basis of insufficient sleep subtypes\u003c/h3\u003e\n\u003cp\u003eStudies using PET and SPECT measured a range of receptors and transporters highly correlated with psychiatric disorders. This enabled us to correlate the spatial patterns of CT in each subtype and the density maps of neurotransmitter and transporter systems (5-hydroxytryptamine: 5HT1a, 5HT1b, 5HT2, and 5HT4; dopamine: D1, D2, and DAT; GABA: GABAa). We found the postcentral subtype (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eg) was significantly correlated with D2, DAT that involved in dopaminergic signaling (r=-0.428, 0.377, respectively, all P\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The pericalcarine subtype (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ei) was significantly correlated with a serotonin receptor 5HT1a (r\u0026thinsp;=\u0026thinsp;0.33, p\u0026thinsp;=\u0026thinsp;0.001). The entorhinal subtype (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eg) was negatively correlated with 5HT1a (r=-0.613), and positively correlated with D2 and GABAa (r\u0026thinsp;=\u0026thinsp;0.507, 0.511 respectively) (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eh).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidation of the subtyping results\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo test the robustness of subtyping results against different definitions of sleep insufficiency, we applied the SuStaIn model to other insufficient sleep cutoffs (7.5 and 8.5 hours) and compared the sequences of biomarker progression with the primary results using the 8 hours cut off (Fig. S3). The SuStaIn trajectories of both replication cutoffs showed high consistency with the original cutoff, with the correlation coefficients (\u003cem\u003er\u003c/em\u003e) of 0.785, 0.811, and 0.689 for the three subtypes between 8 hours and 7.5 hours cutoffs, respectively (P\u003csub\u003eone\u0026minus;tailed\u003c/sub\u003e \u0026lt; 0.001, FDR corrected, Fig. S3e) and \u003cem\u003er\u003c/em\u003e of 0.857, 0.921, and 0.811 for the three subtypes between 8 hours and 8.5 hours cutoffs (P\u003csub\u003eone\u0026minus;tailed\u003c/sub\u003e \u0026lt; 0.001, FDR corrected; Fig. S3f). We further validated the results with subjective sleep cutoff (\u0026lt;\u0026thinsp;8 hours of sleep), which also showed a high consistency (Fig. S6).\u003c/p\u003e\n\u003cp\u003eWe then validated the results on our private dataset (Shanghai Sleep Birth Cohort, SSBC) using the identical procedure. The SuStaIn subtypes found in the SSBC dataset showed high consistency with the discovery set of the ABCD dataset, e.g., subtypes showed starting epicenters in the postcentral cortex (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ea), pericalcarine cortex (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eb), and entorhinal cortex (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ec). For the SSBC dataset, the postcentral subtype was found with similar pediatric sleep presentations as the short sleeper found in ABCD; while the pericalcarine and entorhinal subtype were found of worse pediatric sleep presentations compared with the control group (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ed). The stage progression in the entorhinal subtype was found to be significantly correlated with externalizing problem and the overall problem scores (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ee-g). Such results were also highly in accordance with results from the ABCD study. However, no significant correlation was found between internalizing problems and SuStaIn stage.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eInsufficient sleep has a long-term and heterogeneous impact on the physical and mental health of adolescents\u003csup\u003e41\u003c/sup\u003e, but little is known about its neurobiological mechanisms\u003csup\u003e42\u003c/sup\u003e, especially its progressive and accumulative effect on brain development \u003csup\u003e24,43\u003c/sup\u003e. Previous studies and guidelines for adolescent sleep recommend more than 8 hours of sleep to maximize health and well-being\u003csup\u003e33,34\u003c/sup\u003e. Although sleep duration recommendations inform public policies, medical guidelines and interventions, individual sleep requirements\u003csup\u003e44,45\u003c/sup\u003e and the cause of insufficient sleep duration\u003csup\u003e35\u003c/sup\u003e vary. In this work, we used a data-driven approach (SuStaIn) on a large adolescent sample with both neuroimaging and Fitbit measurement, and identified three subtypes of insufficient sleep based on their cortical thickness, which were associated with distinct socio-ecological factors. Compared with the controls, the pericalcarine subtype was under adverse environmental conditions, along with worse sleep qualities. The entorhinal subtype had a high chance of comorbidity with mental disorder-related behavioral problems, worse sleep qualities, and younger brain age, and was significantly associated with. The postcentral subtype did not show impaired sleep qualities or any comorbidity and showed more advanced development in terms of brain age. Furthermore, the robustness of the three subtypes and their staging was validated with longitudinal data in ABCD, different sleep duration cutoffs, and an independent dataset. These findings emphasize the clinical potential of using brain morphological features for stratification of sleep deficits, as the distinct socio-ecological determinants of different subtypes indicated entirely different interventions.\u003c/p\u003e \u003cp\u003eOne interesting finding in this study is that we identified a subtype of natural short sleepers. It is known that biological demand for total sleep duration is highly personalized\u003csup\u003e17\u003c/sup\u003e. Previous studies also pointed out that there is a group of natural short sleepers who are satisfied with extremely less sleep time, based on genetics, sleep, and other mental or physical measurements.\u003csup\u003e17,46\u003c/sup\u003e In comparison with the control group, despite significantly lower sleeping hours (7.46 +/- 0.51 hours), this postcentral subtype had similar presentations in all sleep-related socio-ecological factors and no negative psychiatric outcomes. Moreover, the postcentral subtype had a significantly higher PRS score for short sleep duration genes, indicating a genetic predisposition for short sleep duration. We also found the shorter sleepers had relatively fewer night wakes in both the ABCD and SSBC datasets (p = 0.051 for the ABCD study and p = 0.161 for the Shanghai dataset; marginally significant), indicating higher sleep efficiency in these adolescents. In the brain development perspective, we found the parietal cortex in this subtype was thinner and reduced faster longitudinally than the controls. Previous studies have found significant correlations between parietal cortex and physical activities in structural and functional domains\u003csup\u003e47–49\u003c/sup\u003e. Compared with other sleep subtypes, the natural short sleepers had a relatively higher physical activity, which may explain the morphological change of the parietal cortex and help to mitigate the negative effects of sleep loss. This speculation was supported by the “brain age”, as the short sleepers were found to have a more advanced brain age compared with the controls, which indicated a significantly higher brain maturity and fast development. As the normative sleep duration decreases during adolescence\u003csup\u003e50,51\u003c/sup\u003e, the more advanced development in this subgroup suggested a natural need for less sleep time. Therefore, it is important to identify this population of nature short sleepers to avoid unnecessary concerns or interventional plans.\u003c/p\u003e \u003cp\u003eFor the other two groups, we identified various external and internal stimulants that may induce sleep deficits and further impair brain development. Sleep environment, including ambient light, noise, and neighborhood depravity, significantly affects adolescent sleep, leading to negative mental and physical consequences\u003csup\u003e52,53\u003c/sup\u003e. A missing link in this causal relationship is how negative sleep environments change the brain. Preclinical models demonstrated that sleep deprivation during adolescence disrupts synaptic pruning\u003csup\u003e54\u003c/sup\u003e—a proposed mechanism for cortical thinning\u003csup\u003e55\u003c/sup\u003e. Among the three SuStaIn subtypes, the pericalcarine subtype was found with worse sleep environmental conditions. As part of the visual network, the pericalcarine cortex exhibits heightened sensitivity to sleep-environment factors. Previous studies found that REM sleep loss (associated with light fragmentation) correlated with reduced default mode network-visual cortex connectivity\u003csup\u003e56\u003c/sup\u003e. We further found that sleep duration mediated ambient light and pericalcarine cortical thickness. These results support the light pollution mechanism proposed by LeGates et al.\u003csup\u003e57\u003c/sup\u003e that light pollution disrupts circadian rhythms and sleep via ipRGCs, which may contribute to brain morphological changes by chronically impairing neuroplasticity.\u003c/p\u003e \u003cp\u003eBesides the external stimulants, many psychiatric disorders are related to sleep deficits\u003csup\u003e7,11,19,58\u003c/sup\u003e. A number of studies have demonstrated that the emergence of psychiatric disorders is partially related to morphological changes caused by abnormal synaptic pruning and myelination\u003csup\u003e59–61\u003c/sup\u003e. For the entorhinal subtype, we found a strong association with psychiatric comorbidity, including both internalizing and externalizing problems. The entorhinal cortex plays an important role in memory processing, mediating cortical information from the hippocampus\u003csup\u003e62\u003c/sup\u003e. Acting as a “gateway” between the hippocampus and neocortex\u003csup\u003e63\u003c/sup\u003e, abnormal thinning of the entorhinal cortex may reflect the impaired cognitive functions\u003csup\u003e64\u003c/sup\u003e due to the comorbidity. During adolescence, developmental changes in neurochemical systems, including dopaminergic and GABA, reform adolescent motivation behaviors, leading to potential internalizing and externalizing disorder appearances\u003csup\u003e60\u003c/sup\u003e. Further analysis revealed that the entorhinal subtype had a much stronger spatial correlation with these psychiatric-related neurotransmitters. The entorhinal subtype was also found to be significantly younger in brain age, suggesting a delay in brain development and indicating potential correlations with psychiatric disorders\u003csup\u003e65–67\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAccording to the layers of sleep socio-ecological factors\u003csup\u003e35,36\u003c/sup\u003e, short sleep duration is associated with environmental factors, physical or mental conditions, and pediatric sleep health issues. The three SuStaIn subtypes are distinctively related to different aspects of sleep socio-ecological factors, and the findings were highly robust, as validated with ABCD longitudinal data and an independent dataset. Several issues should be taken into account for our findings. Firstly, the SuStaIn models pseudo-longitudinal sequences via cross-sectional data, which do not directly reflect the actual progression. Although longitudinal data supported SuStaIn trajectories (ΔCT patterns reflected the SuStaIn subtype trajectories), future work needs to verify the sleep-induced brain alteration trajectories. Secondly, as sleep health is a multidimensional construct, future studies can incorporate more sleep-related variables, including sleep timing and sleep efficiency, to have a more comprehensive stratification of sleep. Moreover, our present research focused on the effects of insufficient sleep on cortical thickness; further studies could incorporate other imaging features and genomic data. In addition, while clustering or subtyping can facilitate stratified interventions, individual differences within clusters still exist. Therefore, dimensional approaches\u003csup\u003e68\u003c/sup\u003e to personalization may further address the individual sleep variations. Thirdly, although the objective sleep measurements in the ABCD study (with Fitbit) and the SSBC dataset (with actigraphy) were measured differently, both Fitbit and actigraphy are accurate measurement tools for sleep duration\u003csup\u003e69\u003c/sup\u003e. Validation experiments with both subjective and objective measurements of sleep duration and cutoffs also support the robustness of insufficient subtype results. Finally, the subjects of PET or SPECT neurochemical density maps were all adults, while our study subjects were adolescents. Further lifespan works could fill this gap towards a better understanding of the adolescent population.\u003c/p\u003e \u003cp\u003eIn summary, we used a data-driven method to disentangle the heterogeneity of the impact of insufficient sleep on the brain, identifying three subtypes with distinct and replicable neuroimaging patterns. This was followed by a comprehensive analysis that revealed the group differences in biological, mental, and environmental determinants. Such results may offer a comprehensive understanding of insufficient sleep on adolescents, calling for stratified attention and interventions towards this social problem.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Method","content":"\u003ch2\u003eData preparation\u003c/h2\u003e\u003cp\u003eIn the second-year follow-up of the ABCD study, 4,134 adolescents with Fitbit worn on weekdays and weekends and with at least seven nights of recorded sleep were included\u003csup\u003e70\u003c/sup\u003e. Note that the baseline ABCD data do not have Fitbit measurements, and thus was not used for discovery. Subjects missing imaging data or covariates (age, sex, BMI, site, race, ethnicity, socioeconomic status) were excluded, leaving 3,266 individuals for formal analysis. The participant inclusion flowchart was provided in Supplementary Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. To derive an average weekday-weekend sleep profile, we calculated the sleep hours weighted according to the number of days [(∑weekday sleep duration) + (∑weekend sleep duration)]/(weekday day counts + weekend day counts) and winsorized\u003csup\u003e21\u003c/sup\u003e. Participants were divided into sufficient or insufficient sleep groups based on a cutoff of 8 hours\u003csup\u003e3,50\u003c/sup\u003e of average sleep based on Fitbit, resulting in 853 adolescents with insufficient sleep and 2,413 with sufficient sleep (age 12.03 ± 0.63 years, 56.5% male for insufficient sleep; age 11.87 ± 0.64 years, 50.6% male for sufficient sleep). We also tested other choices of 7.5 hours and 8.5 hours as cutoffs (Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). The demographic characteristics are summarized in Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003eA private dataset of the Shanghai Sleep Birth Cohort\u003csup\u003e71\u003c/sup\u003e (SSBC dataset) was used for the replication test. This dataset included 52 participants with insufficient sleep and 63 with sufficient sleep based on actigraphy (age 10.33 ± 0.27, 51.9% male for insufficient sleep; age 10.33 ± 0.25 years, 38.7% male for sufficient sleep). As subjects from the SSBC dataset were younger, 8.5 hours of sleep was used as the cutoff for insufficient sleep\u003csup\u003e50\u003c/sup\u003e. All study procedures received approval from institutional review boards at the data collecting sites, and written consent was obtained from parents, with verbal assent from the children.\u003c/p\u003e\u003ch2\u003eImage pre-processing\u003c/h2\u003e\u003cp\u003eWe used the 3D T1-weighted MRI scans from each dataset. The ABCD data were preprocessed by a standard pipeline of cortical reconstruction and volumetric segmentation using FreeSurfer v7.1.1\u003csup\u003e72\u003c/sup\u003e (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://surfer.nmr.mgh.harvard.edu/\u003c/span\u003e\u003cspan address=\"https://surfer.nmr.mgh.harvard.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) from the ABCD study Data Analysis, Informatics, and Resources Center (DAIRC) along with standard quality control measures, reported in previous studies\u003csup\u003e73\u003c/sup\u003e. Aligning with the ABCD study, similar image pre-processing processes were performed for the SSBC dataset. Here we focused on the cortical thickness (CT) features that were shown to be significantly altered in previous studies. We used the Desikan-Killiany Atlas\u003csup\u003e74\u003c/sup\u003e template to parcellate the brain into 68 ROIs. Site-specific variations were estimated and regressed with Neuro-Combat\u003csup\u003e75\u003c/sup\u003e. Before comparing the CT measures between sufficient and insufficient sleep groups, propensity score matching was used to reduce bias caused by confounding variables. Specifically, covariates including age, sex, site, race, ethnicity, socioeconomic status, and BMI were controlled.\u003c/p\u003e\u003ch2\u003ePediatric sleep measurements\u003c/h2\u003e\u003cp\u003eSleep disturbance in ABCD participants is measured based on the Sleep Disturbance Scale for Children (SDSC)\u003csup\u003e76\u003c/sup\u003e with 26 questions, assessing six sleep disturbance dimensions, including Disorders of initiating and maintaining sleep, Sleep breathing disorders, Disorders of arousal, Sleep-wake transition disorders, Disorders of excessive somnolence and Sleep hyperhydrosis. For the SSBC dataset, the Children’s Sleep Habits Questionnaire (CSHQ)\u003csup\u003e77\u003c/sup\u003e was used to assess sleep behavior in eight aspects, includingBedtime resistance, Sleep onset delay, Sleep duration, Sleep anxiety, Night wakings, Parasomnia, Sleep breathing disorders, and Daytime sleepiness. As adolescents usually address weekday sleep debt in weekends\u003csup\u003e3\u003c/sup\u003e, we used weekly sleep loss ((weekly average sleep duration – workday average sleep duration) / workday count) measurement from the Munich Chronotype Questionnaire\u003csup\u003e78\u003c/sup\u003e for evaluation. The higher the weekly sleep loss measure, the larger the workday sleep debt repaid during freedays.\u003c/p\u003e\u003ch2\u003eSleep environment measurements\u003c/h2\u003e\u003cp\u003eIn the ABCD study, the child’s primary residential address was geocoded, and variables of the American Community Survey (5-year estimates from 2011 to 2015) were linked according to the US census tract\u003csup\u003e79\u003c/sup\u003e. Potentially sleep-affecting neighborhood environmental measures were selected, including night light, income, education, noise, and neighborhood disadvantage\u003csup\u003e18\u003c/sup\u003e.\u003c/p\u003e\u003ch2\u003eMental and physical condition measurements\u003c/h2\u003e\u003cp\u003eFor the ABCD dataset, mental health conditions of internalizing, externalizing, and overall problems were assessed with Parent Child Behavior Checklist Scores (CBCL)\u003csup\u003e80\u003c/sup\u003e. For the SSBC dataset, subjects’ mental conditions of internalizing, externalizing, and overall problems were assessed via the Strengths and Difficulties Questionnaire (SDQ)\u003csup\u003e81,82\u003c/sup\u003e. Daily physical activity is measured with step counts with Fitbit.\u003c/p\u003e\u003ch2\u003eBrain age\u003c/h2\u003e\u003cp\u003eIn order to evaluate the difference in brain age between the subtypes, we used the Connectome-Based Predictive Modeling (CPM) framework\u003csup\u003e83,84\u003c/sup\u003e for estimating brain age based on cortical thickness. Cortical thickness from all 68 regions was standardized while regressing out sex and site. Univariate correlation analysis (Pearson’s \u003cem\u003er\u003c/em\u003e) identified regions significantly associated with chronological age (all ROIs have negative correlations with age, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05, adjusted with Bonferroni correction). We used the age-correlated ROIs to train a linear regression model. With 10-fold cross-validation, the model accuracy was evaluated using Pearson’s correlation coefficient (\u003cem\u003er\u003c/em\u003e = 0.17, \u003cem\u003ep\u003c/em\u003e = 2×10\u003csup\u003e− 22\u003c/sup\u003e) between predicted and chronological age.\u003c/p\u003e\u003ch2\u003eNeurotransmitter profiles\u003c/h2\u003e\u003cp\u003eNeurochemical density profiles across cortical regions were quantified through retrospective analysis of positron emission tomography (PET) and single-photon emission computed tomography (SPECT) data obtained from a cohort of healthy controls\u003csup\u003e85\u003c/sup\u003e. Eight receptors across three different neurotransmitter systems (5-hydroxytryptamine: 5HT1a, 5HT1b, 5HT2a, and 5TH4; dopamine: D1, D2, and DAT; GABA: GABAa; details in Supplement 2) were investigated with JuSpace\u003csup\u003e86\u003c/sup\u003e. These density maps were registered to the Montreal Neurological Institute (MNI) stereotaxic coordinate system and segmented into 68 anatomically defined cortical regions using the Desikan-Killiany (DK) atlas\u003csup\u003e74\u003c/sup\u003e.\u003c/p\u003e\u003ch2\u003ePolygenetic risk score\u003c/h2\u003e\u003cp\u003eWe compute polygenetic risk score (PRS) of short sleep duration using imputed ABCD study genotype data with quality control, based on the largest GWAS data of short sleep duration (n = 446,118) from UK BioBank\u003csup\u003e40\u003c/sup\u003e. For sleep duration measurement in this adult dataset, participants were asked about how many hours of sleep within 24 hours, and participants with less than 7 hours of sleep were deemed as short sleepers. Among ABCD children of European ancestry, we used PRSice-2\u003csup\u003e87\u003c/sup\u003e to calculate individual PRSs by adding SNP alleles with weights of the SNP allele effect size estimated in a previous GWAS (short sleep population GWAS). Parameters for polygenic scoring were a clumped variation of linkage disequilibrium r2 \u0026lt; .10 within a 300-kb window cutoff. We calculated PRSs for each trait based on six different p-value thresholds (pTs) with pT \u0026lt; 5e.08 (32 SNPs), pT \u0026lt; 0.001 (4,009 SNP), pT \u0026lt; 0.005 (12,675 SNPs), pT \u0026lt; 0.01 (22,140 SNPs), pT \u0026lt; 0.05 (73,429 SNPs), pT \u0026lt; 0.1 (129,416 SNPs), and pT \u0026lt; 0.5 (422,712 SNPs), as recommended\u003csup\u003e87\u003c/sup\u003e. An optimal p-value threshold (pT \u0026lt; 0.5) was found with the highest percentage of the variance (R\u003csup\u003e2\u003c/sup\u003e = 0.016, p = 4.45e-05) in the outcome estimated by PRSice-2, and were further Z-score normalized. A higher PRS score indicates a genetic predisposition for short sleep duration.\u003c/p\u003e\u003ch2\u003eSubtyping of insufficient-sleep related brain morphology\u003c/h2\u003e\u003cp\u003eWe employed a widely used data-driven neuroimage morphological progression model named Subtype and Stage Inference (SuStaIn)\u003csup\u003e32\u003c/sup\u003e, which identifies progression subtypes with an event-based model. The SuStaIn model assigns subtypes and stages to individuals based on their biomarkers. The estimated SuStaIn stages were used to capture the progression of diseases, with earlier stages meaning little biomarker changes and later stages meaning severe biomarker changes.\u003c/p\u003e\u003cp\u003eWe first regressed out confounding factors (age, sex, site, race, ethnicity, socioeconomic status, and BMI) from regions of interest (ROIs) using linear regression. Scanner-induced effects were regressed with Neuro-Combat\u003csup\u003e75\u003c/sup\u003e. The adjusted ROIs of the insufficient sleep group were z-scored relative to the sufficient sleep group, with a higher z-value indicating greater deviations from sufficient sleep. We selected twelve ROIs with significantly reduced CT in the insufficient sleep group as SuStaIn inputs (P\u003csub\u003eFDR\u003c/sub\u003e\u0026lt;0.05, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, d). Based on the parameter settings outlined in previous research, we used an expectation-maximization algorithm for model initialization and used 25 random starting points to give the optimum solution. We applied 1,000,000 rounds of Markov Chain Monte Carlo (MCMC) to estimate the developmental trajectories of the identified subtypes. Three z-score cutoffs (1, 1.5, and 2) were employed to describe the amount of negative impact of sleep deficits on adolescents. We assessed subtyping solutions with 2–5 clusters, and an optimal choice of 3 was chosen according to its lower CVIC\u003csup\u003e88\u003c/sup\u003e and higher log-likelihood through ten-fold cross-validation (Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). To test the stability of the results, other z-score cutoffs for SuStaIn were also tested in Supplementary Fig. S4.\u003c/p\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eAfter identifying SuStaIn insufficient sleep subtypes, we compared different socio-ecological factors affecting sleep among the subtypes for their pediatric sleep health, physical or mental conditions, and family and neighborhood factors using the Mann-Whitney U test and computed the effect size using Cohen’s d test.\u003c/p\u003e\u003cp\u003eWe further compared the brain age (in measurement for brain maturation) and short sleep PRS (in measurement for genetic backgrounds) among subtypes using the Mann-Whitney U test. As for the associations among night light pollution, sleep duration, and brain structure, we performed mediation analysis using sleep duration as the mediator. All variables were normalized before entering the model. Sex, age, parent education level, household income, ethnicity, BMI, and site were used as covariates. Total, direct, and indirect associations were quantified through 10,000 bootstrap iterations, with 95% bias-corrected and accelerated confidence intervals calculated. All statistical analyses were performed using the R mediation package (v4.0.0). The statistical significance threshold for these analyses was set at P\u003csub\u003eBonferroni\u003c/sub\u003e \u0026lt; 0.05. To understand the subtype-specific neurochemical signatures, we performed a spatial correlation using bootstrapped Spearman correlation with 10,000 iterations between each neurochemical density matrix and cortical thickness T-statistic maps for each subtype.\u003c/p\u003e\u003ch2\u003eLongitudinal validation\u003c/h2\u003e\u003cp\u003eTo test the reliability of the SuStaIn staging resultswe utilized ABCD longitudinal data at the first and the second visits, which were two years apart. We calculated the actual change of cortical thickness of individuals in comparison with the pseudo-progression of SuStaIn. The change of cortical thickness (ΔCT) was obtained by subtracting the value of the baseline year from the second-year follow-up in each ROI while regressing the age gap between the scans, which were compared between the sufficient and insufficient sleep groups.\u003c/p\u003e\u003ch2\u003eIndependent validation\u003c/h2\u003e\u003cp\u003eWe applied SustaIn model train via the ABCD sample on a private dataset of the Shanghai Sleep Birth Cohort. We compared the brain phenotype, pediatric sleep, and psychiatric comorbid between the identified subtypes and control. The results showed consistency.\u003c/p\u003e\u003cp\u003eFurthermore, to test the robustness of subtyping results against different definitions of sleep insufficiency, we retrained the SustaIn on other insufficient sleep objective sleep cutoffs (7.5 and 8.5 hours). As sleep duration is also measured via structured scales, we used the Sleep Disturbance Scale for Children (SDSC) measured sleep duration as a cutoff (\u0026lt; 8 hours of sleep), and retrained the SustaIn model. We compared the consistency of results obtained by different cutoff criteria via linear correlation between the original and the retrained sequences of biomarker progression.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data repository houses all data generated by the Adolescent Brain Cognitive Development (ABCD) Study at https://nda.nih.gov/abcd/. Data will be available upon reasonable request to the corresponding author after institutional approval and with a signed data access agreement or with the permission of the Shanghai Children\u0026apos;s Medical Center.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePython of the SuStaIn algorithm is available on\u0026nbsp;https://github.com/ucl-pond.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDuring the study, Prof. Dan Wu is supported by Ministry of Science and Technology of the People\u0026rsquo;s Republic of China (2018YFE0114600, 2021ZD0200202), National Natural Science Foundation of China (81971606, 82122032), and Science and Technology Department of Zhejiang Province (202006140, 2022C03057).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eProf. Guanghai Wang is supported by the National Science and Technology Innovation 2030 Major Project of China (STI2030-Major Projects+2021ZD0204200), National Natural Science Foundation of China (82071493, 82073568), Shanghai Municipal Health Commission (2022XD056), and Innovative Research Team of High-level Local Universities in Shanghai (SHSMU-ZDCX20211100).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe authors declare no competing interests.\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAltman NG, Izci-Balserak B, Schopfer E, et al. Sleep duration versus sleep insufficiency as predictors of cardiometabolic health outcomes. Sleep Medicine. 2012;13(10):1261-1270. doi:10.1016/j.sleep.2012.08.005\u003c/li\u003e\n\u003cli\u003eGradisar M, Gardner G, Dohnt H. Recent worldwide sleep patterns and problems during adolescence: A review and meta-analysis of age, region, and sleep. Sleep Medicine. 2011;12(2):110-118. doi:10.1016/j.sleep.2010.11.008\u003c/li\u003e\n\u003cli\u003eOwens J, ADOLESCENT SLEEP WORKING GROUP, COMMITTEE ON ADOLESCENCE, et al. Insufficient Sleep in Adolescents and Young Adults: An Update on Causes and Consequences. Pediatrics. 2014;134(3):e921-e932. doi:10.1542/peds.2014-1696\u003c/li\u003e\n\u003cli\u003eLim DC, Najafi A, Afifi L, et al. The need to promote sleep health in public health agendas across the globe. The Lancet Public Health. 2023;8(10):e820-e826. doi:10.1016/S2468-2667(23)00182-2\u003c/li\u003e\n\u003cli\u003eKirshenbaum JS, Coury SM, Colich NL, Manber R, Gotlib IH. Objective and subjective sleep health in adolescence: Associations with puberty and affect. Journal of Sleep Research. 2023;32(3):e13805. doi:10.1111/jsr.13805\u003c/li\u003e\n\u003cli\u003eYan YW, Lin RM, Su YK, Liu MY. The relationship between adolescent academic stress and sleep quality: A multiple mediation model. soc behav pers. 2018;46(1):63-77. doi:10.2224/sbp.6530\u003c/li\u003e\n\u003cli\u003eFreeman D, Sheaves B, Waite F, Harvey AG, Harrison PJ. Sleep disturbance and psychiatric disorders. The Lancet Psychiatry. 2020;7(7):628-637. doi:10.1016/S2215-0366(20)30136-X\u003c/li\u003e\n\u003cli\u003eTroynikov O, Watson CG, Nawaz N. Sleep environments and sleep physiology: A review. Journal of Thermal Biology. 2018;78:192-203. doi:10.1016/j.jtherbio.2018.09.012\u003c/li\u003e\n\u003cli\u003eKnutson KL, Van Cauter E. Associations between Sleep Loss and Increased Risk of Obesity and Diabetes. Annals of the New York Academy of Sciences. 2008;1129(1):287-304. doi:10.1196/annals.1417.033\u003c/li\u003e\n\u003cli\u003ePrather AA, Leung CW. Association of Insufficient Sleep With Respiratory Infection Among Adults in the United States. JAMA Intern Med. 2016;176(6):850. doi:10.1001/jamainternmed.2016.0787\u003c/li\u003e\n\u003cli\u003eSun X, Liu B, Liu S, et al. Sleep disturbance and psychiatric disorders: a bidirectional Mendelian randomisation study. Epidemiol Psychiatr Sci. 2022;31:e26. doi:10.1017/S2045796021000810\u003c/li\u003e\n\u003cli\u003eGordon AM, Carrillo B, Barnes CM. Sleep and social relationships in healthy populations: A systematic review. Sleep Medicine Reviews. 2021;57:101428. doi:10.1016/j.smrv.2021.101428\u003c/li\u003e\n\u003cli\u003eMah CD, Mah KE, Kezirian EJ, Dement WC. The Effects of Sleep Extension on the Athletic Performance of Collegiate Basketball Players. Sleep. 2011;34(7):943-950. doi:10.5665/SLEEP.1132\u003c/li\u003e\n\u003cli\u003eJackson ML, Gunzelmann G, Whitney P, et al. Deconstructing and reconstructing cognitive performance in sleep deprivation. Sleep Medicine Reviews. 2013;17(3):215-225. doi:10.1016/j.smrv.2012.06.007\u003c/li\u003e\n\u003cli\u003eHill ED, Kashyap P, Raffanello E, et al. Prediction of mental health risk in adolescents. Nat Med. Published online March 5, 2025. doi:10.1038/s41591-025-03560-7\u003c/li\u003e\n\u003cli\u003eGrandner MA, Patel NP, Gehrman PR, Perlis ML, Pack AI. Problems associated with short sleep: Bridging the gap between laboratory and epidemiological studies. Sleep Medicine Reviews. 2010;14(4):239-247. doi:10.1016/j.smrv.2009.08.001\u003c/li\u003e\n\u003cli\u003eHe Y, Jones CR, Fujiki N, et al. The Transcriptional Repressor DEC2 Regulates Sleep Length in Mammals. Science. 2009;325(5942):866-870. doi:10.1126/science.1174443\u003c/li\u003e\n\u003cli\u003eMayne SL, Mitchell JA, Virudachalam S, Fiks AG, Williamson AA. Neighborhood environments and sleep among children and adolescents: A systematic review. Sleep Medicine Reviews. 2021;57:101465. doi:10.1016/j.smrv.2021.101465\u003c/li\u003e\n\u003cli\u003eOhi K, Ochi R, Noda Y, et al. Polygenic risk scores for major psychiatric and neurodevelopmental disorders contribute to sleep disturbance in childhood: Adolescent Brain Cognitive Development (ABCD) Study. Transl Psychiatry. 2021;11(1):187. doi:10.1038/s41398-021-01308-8\u003c/li\u003e\n\u003cli\u003eCooper R, Di Biase MA, Bei B, Quach J, Cropley V. Associations of Changes in Sleep and Emotional and Behavioral Problems From Late Childhood to Early Adolescence. JAMA Psychiatry. 2023;80(6):585. doi:10.1001/jamapsychiatry.2023.0379\u003c/li\u003e\n\u003cli\u003eZhang L, Sasser J, Doane LD, Peltz J, Oshri A. Latent Profiles of Sleep Patterns in Early Adolescence: Associations With Behavioral Health Risk. Journal of Adolescent Health. 2024;74(1):177-185. doi:10.1016/j.jadohealth.2023.08.021\u003c/li\u003e\n\u003cli\u003eSehgal A, Mignot E. Genetics of Sleep and Sleep Disorders. Cell. 2011;146(2):194-207. doi:10.1016/j.cell.2011.07.004\u003c/li\u003e\n\u003cli\u003eJha PK, Valekunja UK, Ray S, Nollet M, Reddy AB. Single-cell transcriptomics and cell-specific proteomics reveals molecular signatures of sleep. Commun Biol. 2022;5(1):846. doi:10.1038/s42003-022-03800-3\u003c/li\u003e\n\u003cli\u003eKrause AJ. The sleep-deprived human brain.\u003c/li\u003e\n\u003cli\u003eCheng W, Rolls E, Gong W, et al. Sleep duration, brain structure, and psychiatric and cognitive problems in children. Mol Psychiatry. 2021;26(8):3992-4003. doi:10.1038/s41380-020-0663-2\u003c/li\u003e\n\u003cli\u003eYang FN, Xie W, Wang Z. Effects of sleep duration on neurocognitive development in early adolescents in the USA: a propensity score matched, longitudinal, observational study. The Lancet Child \u0026amp; Adolescent Health. 2022;6(10):705-712. doi:10.1016/S2352-4642(22)00188-2\u003c/li\u003e\n\u003cli\u003eJalbrzikowski M, Hayes RA, Scully KE, et al. Associations between brain structure and sleep patterns across adolescent development. Sleep. 2021;44(10):zsab120. doi:10.1093/sleep/zsab120\u003c/li\u003e\n\u003cli\u003eDutil C, Walsh JJ, Featherstone RB, et al. Influence of sleep on developing brain functions and structures in children and adolescents: A systematic review. Sleep Medicine Reviews. 2018;42:184-201. doi:10.1016/j.smrv.2018.08.003\u003c/li\u003e\n\u003cli\u003eWang Y, Genon S, Dong D, et al. Covariance patterns between sleep health domains and distributed intrinsic functional connectivity. Nat Commun. 2023;14(1):7133. doi:10.1038/s41467-023-42945-5\u003c/li\u003e\n\u003cli\u003eMa Q, Sahakian BJ, Zhang B, Li Z. Neural correlates of device-based sleep characteristics in adolescents.\u003c/li\u003e\n\u003cli\u003eBresser T, Blanken TF, De Lange SC, et al. Insomnia Subtypes Have Differentiating Deviations in Brain Structural Connectivity. Biological Psychiatry. 2025;97(3):302-312. doi:10.1016/j.biopsych.2024.06.014\u003c/li\u003e\n\u003cli\u003eYoung AL. Uncovering the heterogeneity and temporal complexity of neurodegenerative diseases with Subtype and Stage Inference. NATURE COMMUNICATIONS. Published online 2018.\u003c/li\u003e\n\u003cli\u003eHirshkowitz M, Whiton K, Albert SM, et al. National Sleep Foundation\u0026rsquo;s updated sleep duration recommendations: final report. Sleep Health. 2015;1(4):233-243. doi:10.1016/j.sleh.2015.10.004\u003c/li\u003e\n\u003cli\u003eParuthi S, Brooks LJ, D\u0026rsquo;Ambrosio C, et al. Recommended Amount of Sleep for Pediatric Populations: A Consensus Statement of the American Academy of Sleep Medicine. Journal of Clinical Sleep Medicine. 2016;12(06):785-786. doi:10.5664/jcsm.5866\u003c/li\u003e\n\u003cli\u003eKwon M, Page SD, Williamson AA, Morgan S, Sawyer AM. Social determinants of health at multiple socio-ecological levels and sleep health in adolescents: A scoping review. Sleep Medicine Reviews. 2024;78:102008. doi:10.1016/j.smrv.2024.102008\u003c/li\u003e\n\u003cli\u003eMeltzer LJ, Williamson AA, Mindell JA. Pediatric sleep health: It matters, and so does how we define it. Sleep Medicine Reviews. 2021;57:101425. doi:10.1016/j.smrv.2021.101425\u003c/li\u003e\n\u003cli\u003eNewton AT, Honaker SM, Reid GJ. Risk and protective factors and processes for behavioral sleep problems among preschool and early school-aged children: A systematic review. Sleep Medicine Reviews. 2020;52:101303. doi:10.1016/j.smrv.2020.101303\u003c/li\u003e\n\u003cli\u003eTarokh L. Sleep in adolescence: Physiology, cognition and mental health. Neuroscience and Biobehavioral Reviews. Published online 2016.\u003c/li\u003e\n\u003cli\u003eHolm MC, Leonardsen EH, Beck D, et al. Linking brain maturation and puberty during early adolescence using longitudinal brain age prediction in the ABCD cohort. Developmental Cognitive Neuroscience. 2023;60:101220. doi:10.1016/j.dcn.2023.101220\u003c/li\u003e\n\u003cli\u003eDashti HS, Jones SE, Wood AR, et al. Genome-wide association study identifies genetic loci for self-reported habitual sleep duration supported by accelerometer-derived estimates. Nat Commun. 2019;10(1):1100. doi:10.1038/s41467-019-08917-4\u003c/li\u003e\n\u003cli\u003eSadeh A. Consequences of Sleep Loss or Sleep Disruption in Children. Sleep Medicine Clinics. 2007;2(3):513-520. doi:10.1016/j.jsmc.2007.05.012\u003c/li\u003e\n\u003cli\u003eEmbang JEG, Tan YHV, Ng YX, et al. Role of sleep and neurochemical biomarkers in synaptic plasticity related to neurological and psychiatric disorders: A scoping review.\u003c/li\u003e\n\u003cli\u003eMedic G, Wille M, Hemels ME. Short- and long-term health consequences of sleep disruption. Nature and Science of Sleep.\u003c/li\u003e\n\u003cli\u003eMatricciani L, Blunden S, Rigney G, Williams MT, Olds TS. Children\u0026rsquo;s Sleep Needs: Is There Sufficient Evidence to Recommend Optimal Sleep for Children? Sleep. 2013;36(4):527-534. doi:10.5665/sleep.2538\u003c/li\u003e\n\u003cli\u003eFeinberg I. Recommended Sleep Durations for Children and Adolescents: The Dearth of Empirical Evidence. Sleep. 2013;36(4):461-462. doi:10.5665/sleep.2520\u003c/li\u003e\n\u003cli\u003eAshbrook LH. Genetics of the human circadian clock and sleep homeostat. Published online 2020.\u003c/li\u003e\n\u003cli\u003eGomes Da Silva S, Arida RM. Physical activity and brain development. Expert Review of Neurotherapeutics. 2015;15(9):1041-1051. doi:10.1586/14737175.2015.1077115\u003c/li\u003e\n\u003cli\u003eValkenborghs SR, Noetel M, Hillman CH, et al. The Impact of Physical Activity on Brain Structure and Function in Youth: A Systematic Review. Pediatrics. 2019;144(4):e20184032. doi:10.1542/peds.2018-4032\u003c/li\u003e\n\u003cli\u003eTuan LH, Tsao CY, Lee LJH, Lee LJ. Voluntary exercise ameliorates synaptic pruning deficits in sleep-deprived adolescent mice. Brain, Behavior, and Immunity. 2021;93:96-110. doi:10.1016/j.bbi.2020.12.017\u003c/li\u003e\n\u003cli\u003eGalland BC, Short MA, Terrill P, et al. Establishing normal values for pediatric nighttime sleep measured by actigraphy: a systematic review and meta-analysis. Sleep. 2018;41(4). doi:10.1093/sleep/zsy017\u003c/li\u003e\n\u003cli\u003eOlds T, Maher C, Blunden S, Matricciani L. Normative Data on the Sleep Habits of Australian Children and Adolescents. 2010;33(10).\u003c/li\u003e\n\u003cli\u003eZeringue MM, Saini EK, Fuller-Rowell TE, Hinnant JB, El-Sheikh M. Neighborhood environment and adolescent sleep: The role of family socioeconomic status. Sleep Medicine. 2023;109:40-49. doi:10.1016/j.sleep.2023.06.014\u003c/li\u003e\n\u003cli\u003eHunter JC, Hayden KM. The association of sleep with neighborhood physical and social environment. Public Health. 2018;162:126-134. doi:10.1016/j.puhe.2018.05.003\u003c/li\u003e\n\u003cli\u003eMaret S, Faraguna U, Nelson AB, Cirelli C, Tononi G. Sleep and waking modulate spine turnover in the adolescent mouse cortex. Nat Neurosci. 2011;14(11):1418-1420. doi:10.1038/nn.2934\u003c/li\u003e\n\u003cli\u003eHuttenlocher PR, Dabholkar AS. Regional differences in synaptogenesis in human cerebral cortex. Journal of Comparative Neurology. 1997;387(2):167-178. doi:10.1002/(SICI)1096-9861(19971020)387:2\u0026lt;167::AID-CNE1\u0026gt;3.0.CO;2-Z\u003c/li\u003e\n\u003cli\u003eDi T, Zhang L, Meng S, et al. The impact of REM sleep loss on human brain connectivity. Transl Psychiatry. 2024;14(1):270. doi:10.1038/s41398-024-02985-x\u003c/li\u003e\n\u003cli\u003eLeGates TA, Fernandez DC, Hattar S. Light as a central modulator of circadian rhythms, sleep and affect. Nat Rev Neurosci. 2014;15(7):443-454. doi:10.1038/nrn3743\u003c/li\u003e\n\u003cli\u003eLi Y, Gong W, Sahakian BJ, et al. Divergent biological pathways linking short and long sleep durations to mental and physical health. Nat Mental Health. 2025;3(4):429-443. doi:10.1038/s44220-025-00395-6\u003c/li\u003e\n\u003cli\u003eKeshavan MS, Giedd J, Lau JYF, Lewis DA, Paus T. Changes in the adolescent brain and the pathophysiology of psychotic disorders. Lancet Psychiatry. 2014;1(7):549-558. doi:10.1016/S2215-0366(14)00081-9\u003c/li\u003e\n\u003cli\u003ePaus T, Keshavan M, Giedd JN. Why do many psychiatric disorders emerge during adolescence? Nat Rev Neurosci. 2008;9(12):947-957. doi:10.1038/nrn2513\u003c/li\u003e\n\u003cli\u003eSolmi M, Radua J, Olivola M, et al. Age at onset of mental disorders worldwide: large-scale meta-analysis of 192 epidemiological studies. Mol Psychiatry. 2022;27(1):281-295. doi:10.1038/s41380-021-01161-7\u003c/li\u003e\n\u003cli\u003eGarcia AD, Buffalo EA. Anatomy and Function of the Primate Entorhinal Cortex. Annu Rev Vis Sci. 2020;6:411-432. doi:10.1146/annurev-vision-030320-041115\u003c/li\u003e\n\u003cli\u003eTakehara-Nishiuchi K, Maal-Bared G, Morrissey MD. Increased Entorhinal\u0026ndash;Prefrontal Theta Synchronization Parallels Decreased Entorhinal\u0026ndash;Hippocampal Theta Synchronization during Learning and Consolidation of Associative Memory. Front Behav Neurosci. 2012;5:90. doi:10.3389/fnbeh.2011.00090\u003c/li\u003e\n\u003cli\u003eKnopman DS, Lundt ES, Therneau TM, et al. Entorhinal cortex tau, amyloid-\u0026beta;, cortical thickness and memory performance in non-demented subjects. Brain. 2019;142(4):1148-1160. doi:10.1093/brain/awz025\u003c/li\u003e\n\u003cli\u003eSchmaal L, Hibar DP, S\u0026auml;mann PG, et al. Cortical abnormalities in adults and adolescents with major depression based on brain scans from 20 cohorts worldwide in the ENIGMA Major Depressive Disorder Working Group. Mol Psychiatry. 2017;22(6):900-909. doi:10.1038/mp.2016.60\u003c/li\u003e\n\u003cli\u003eShaw P, Eckstrand K, Sharp W, et al. Attention-deficit/hyperactivity disorder is characterized by a delay in cortical maturation. Proceedings of the National Academy of Sciences. 2007;104(49):19649-19654. doi:10.1073/pnas.0707741104\u003c/li\u003e\n\u003cli\u003eLuby JL, Belden AC, Jackson JJ, et al. Early Childhood Depression and Alterations in the Trajectory of Gray Matter Maturation in Middle Childhood and Early Adolescence. JAMA Psychiatry. 2016;73(1):31-38. doi:10.1001/jamapsychiatry.2015.2356\u003c/li\u003e\n\u003cli\u003eMeyer N, Harvey AG, Lockley SW, Dijk DJ. Circadian rhythms and disorders of the timing of sleep. The Lancet. 2022;400(10357):1061-1078. doi:10.1016/S0140-6736(22)00877-7\u003c/li\u003e\n\u003cli\u003eEylon G, Tikotzky L, Dinstein I. Performance evaluation of Fitbit Charge 3 and actigraphy vs. polysomnography: Sensitivity, specificity, and reliability across participants and nights. Sleep Health. 2023;9(4):407-416. doi:10.1016/j.sleh.2023.04.001\u003c/li\u003e\n\u003cli\u003eAili K, \u0026Aring;str\u0026ouml;m-Paulsson S, Stoetzer U, Svartengren M, Hillert L. Reliability of Actigraphy and Subjective Sleep Measurements in Adults: The Design of Sleep Assessments. Journal of Clinical Sleep Medicine. 2017;13(01):39-47. doi:10.5664/jcsm.6384\u003c/li\u003e\n\u003cli\u003eLin J, Sun W, Song Y, et al. Cohort Profile: The Shanghai Sleep Birth Cohort Study. Paediatric Perinatal Epid. 2021;35(2):257-268. doi:10.1111/ppe.12738\u003c/li\u003e\n\u003cli\u003eDale AM, Fischl B, Sereno MI. Cortical Surface-Based Analysis: I. Segmentation and Surface Reconstruction. NeuroImage. 1999;9(2):179-194. doi:10.1006/nimg.1998.0395\u003c/li\u003e\n\u003cli\u003eCasey BJ, Cannonier T, Conley MI, et al. The Adolescent Brain Cognitive Development (ABCD) study: Imaging acquisition across 21 sites. Developmental Cognitive Neuroscience. 2018;32:43-54. doi:10.1016/j.dcn.2018.03.001\u003c/li\u003e\n\u003cli\u003eDesikan RS, Segonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Published online 2006.\u003c/li\u003e\n\u003cli\u003ePomponio R, Erus G, Habes M, et al. Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan. NeuroImage. 2020;208:116450. doi:10.1016/j.neuroimage.2019.116450\u003c/li\u003e\n\u003cli\u003eBruni O, Ottaviano S, Guidetti V, et al. The Sleep Disturbance Scale for Children (SDSC) Construct ion and validation of an instrument to evaluate sleep disturbances in childhood and adolescence. Journal of Sleep Research. 1996;5(4):251-261. doi:10.1111/j.1365-2869.1996.00251.x\u003c/li\u003e\n\u003cli\u003eOwens JA, Spirito A, McGuinn M. The Children\u0026rsquo;s Sleep Habits Questionnaire (CSHQ): Psychometric Properties of A Survey Instrument for School-Aged Children. Sleep. 2000;23(8):1-9. doi:10.1093/sleep/23.8.1d\u003c/li\u003e\n\u003cli\u003eZavada A, Gordijn MCM, Beersma DGM, Daan S, Roenneberg T. Comparison of the Munich Chronotype Questionnaire with the Horne‐\u0026Ouml;stberg\u0026rsquo;s Morningness‐Eveningness score. Chronobiology International. 2005;22(2):267-278. doi:10.1081/CBI-200053536\u003c/li\u003e\n\u003cli\u003eCardenas-Iniguez C, Schachner JN, Ip KI, et al. Building towards an adolescent neural urbanome: Expanding environmental measures using linked external data (LED) in the ABCD study. Developmental Cognitive Neuroscience. 2024;65:101338. doi:10.1016/j.dcn.2023.101338\u003c/li\u003e\n\u003cli\u003eAchenbach TM, Ruffle TM. The Child Behavior Checklist and Related Forms for Assessing Behavioral/Emotional Problems and Competencies. Pediatrics In Review. 2000;21(8):265-271. doi:10.1542/pir.21.8.265\u003c/li\u003e\n\u003cli\u003eGoodman R. The Strengths and Difficulties Questionnaire: A Research Note. Child Psychology Psychiatry. 1997;38(5):581-586. doi:10.1111/j.1469-7610.1997.tb01545.x\u003c/li\u003e\n\u003cli\u003eGoodman A, Lamping DL, Ploubidis GB. When to Use Broader Internalising and Externalising Subscales Instead of the Hypothesised Five Subscales on the Strengths and Difficulties Questionnaire (SDQ): Data from British Parents, Teachers and Children. J Abnorm Child Psychol. 2010;38(8):1179-1191. doi:10.1007/s10802-010-9434-x\u003c/li\u003e\n\u003cli\u003eSun H, Mehta S, Khaitova M, et al. Brain age prediction and deviations from normative trajectories in the neonatal connectome. Nat Commun. 2024;15(1):10251. doi:10.1038/s41467-024-54657-5\u003c/li\u003e\n\u003cli\u003eShen X, Finn ES, Scheinost D, et al. Using connectome-based predictive modeling to predict individual behavior from brain connectivity. Nat Protoc. 2017;12(3):506-518. doi:10.1038/nprot.2016.178\u003c/li\u003e\n\u003cli\u003eHansen JY, Shafiei G, Markello RD, et al. Mapping neurotransmitter systems to the structural and functional organization of the human neocortex. Nat Neurosci. 2022;25(11):1569-1581. doi:10.1038/s41593-022-01186-3\u003c/li\u003e\n\u003cli\u003eDukart J, Holiga S, Rullmann M, et al. JuSpace: A tool for spatial correlation analyses of magnetic resonance imaging data with nuclear imaging derived neurotransmitter maps.\u003c/li\u003e\n\u003cli\u003eChoi SW, Mak TSH, O\u0026rsquo;Reilly PF. Tutorial: a guide to performing polygenic risk score analyses. Nat Protoc. 2020;15(9):2759-2772. doi:10.1038/s41596-020-0353-1\u003c/li\u003e\n\u003cli\u003eGelman A, Hwang J, Vehtari A. Understanding predictive information criteria for Bayesian models. Stat Comput. Published online 2014.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6851158/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6851158/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eInsufficient sleep has become an increasing public health issue in adolescents, which is associated with complex social-ecological factors and a wide range of neurodevelopmental outcomes. The highly heterogeneous determinants and outcomes of insufficient sleep impose challenges in effective interventions. To understand such heterogeneity in terms of its impact on the brain, we employed a data-driven ​Subtype and Stage Inference (SuStaIn)​ model to classify the spatiotemporal trajectories of MRI-based neurobiology in 3,266 adolescents from the ABCD study (853 of them had sleep duration less than 8 hours according to Fitbit measurements). We identified three distinct subtypes with reduced cortical thickness, starting from the postcentral cortex, pericalcarine cortex, and entorhinal cortex, respectively. These subtypes diverged significantly in sleep-related social-ecological factors. The postcentral-originated subtype mirrored healthy controls in sleep behavior and sleep environment and showed no psychiatric comorbidities, which aligned phenotypically with natural short sleepers. Notably, this subtype displayed significantly advanced brain age, suggesting advanced neurodevelopment that explained the reduced sleep demand, and polygenetic score analysis revealed a genetic predisposition for short sleep in these adolescents. The pericalcarine-originated subtype displayed environmentally driven sleep insufficiency (e.g., light/noise pollution) where sleep duration mediated environmental effects on pericalcarine cortical thinning. The entorhinal subtype showed elevated psychiatric risk, younger brain age, and spatial correlations with psychosis-related neurotransmitter systems. This work deciphered heterogeneous impacts of insufficient sleep on the brain and the relation with biological, mental, and environmental determinants, offering a framework to guide stratified prevention and intervention strategies.\u003c/p\u003e","manuscriptTitle":"Heterogeneous Neurophenotypes of Adolescent Sleep Insufficiency Stratify Natural Short Sleepers from Comorbidity or Environment Driven Insufficiency","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-25 07:40:58","doi":"10.21203/rs.3.rs-6851158/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"15a8d008-583d-4cb7-9b40-0c87a85d17b9","owner":[],"postedDate":"June 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":50473130,"name":"Health sciences/Health care/Paediatrics/Paediatric research"},{"id":50473131,"name":"Health sciences/Health care/Disease prevention/Lifestyle modification"},{"id":50473132,"name":"Health sciences/Biomarkers/Prognostic markers"}],"tags":[],"updatedAt":"2026-04-21T07:05:35+00:00","versionOfRecord":{"articleIdentity":"rs-6851158","link":"https://doi.org/10.1038/s41467-026-70135-6","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2026-03-07 05:00:00","publishedOnDateReadable":"March 7th, 2026"},"versionCreatedAt":"2025-06-25 07:40:58","video":"","vorDoi":"10.1038/s41467-026-70135-6","vorDoiUrl":"https://doi.org/10.1038/s41467-026-70135-6","workflowStages":[]},"version":"v1","identity":"rs-6851158","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6851158","identity":"rs-6851158","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}