Molecular Autism

Prevalence in autism spectrum disorder (ASD) diagnosis has long been strongly male-biased. Yet, consensus has not been reached on mechanisms and clinical features that underlie sex-based discrepancies. Whereas females may be under-diagnosed because of inconsistencies in diagnostic/ascertainment procedures (sex-biased criteria, social camouflaging), diagnosed males may have exhibited more overt behaviors (e.g., hyperactivity, aggression) that prompted clinical evaluation. Applying a novel network-theory-based approach, we extracted data-driven, clinically-relevant insights from a large, well-characterized sample (Simons Simplex Collection) of 2175 autistic males (Ages = 8.9{+/-}3.5 years) and 334 autistic females (Ages = 9.2{+/-}3.7 years). Exploratory factor analysis (EFA) and expert clinical review reduced data dimensionality to 15 factors of interest. To offset inherent confounds of an imbalanced sample, we identified a subset of males (N=331) matched to females on key variables (Age, IQ) and applied data-driven CDA using Greedy Fast Causal Inference (GFCI) for three groups (All Females, All Males, and Matched Males). Structural equation modeling (SEM) extracted measures of model fit and effect sizes for causal relationships between sex, age, and, IQ on EFA-selected factors capturing phenotypic representations of autism across sensory, social, and restricted and repetitive behavior domains. Our methodology unveiled sex-specific directional relationships to inform developmental outcomes and targeted interventions.

Language development, a core pillar of social communication, has variable trajectories in autism that include a regression or loss of skills in roughly 20% of autistic individuals. Language regression is most frequently identified through parent report but can also be observed as a decrease in raw scores on a repeated language assessment (measure-defined). Later language outcomes after regression have been observed to be highly variable, but not lower than children without a language regression. The current study explores rates of parent-reported and measure-defined language regression in a large sample of infants at high familial likelihood of autism due to having an older autistic sibling. Among all participants at high familial likelihood for autism (n=428), parent-reported regression was observed in 2.8% (n=12) and was associated with 2.77 times higher odds of receiving an autism diagnosis. Measure-defined regression was observed in 8% (n=36) and was associated with 1.21 times higher odds of autism diagnosis. These rates of regression are expectedly lower than estimates collected in autistic samples. Neither of these elevated odds was statistically significant and there was low concordance between these groups with only one participant present in both. Nearest-neighbor comparison samples of non-autistic infants at high and low likelihood for autism without language regression were selected to assess differences in language growth trajectories associated with regression. Infants with parent-reported language regression showed comparable language development to a matched high-likelihood sample while infants with measure-defined language regression showed slower overall language development than matched peers. Taken together, our results show that parent-report and direct measurement of regression capture unique aspects of child language development that may not be predictive of an autism diagnosis but may indicate delayed language growth in early toddlerhood. These language outcomes support previous findings of wide heterogeneity among those with regression and continued language growth after loss of skills. Key PointsO_LILanguage regression can be captured through parent-report or decrease in raw scores on repeated language assessment and is reported in approximately 20% of autistic toddlers. C_LIO_LIMost research on language regression uses retrospective report of regression in autistic children, but this study prospectively examines regression in toddlers at high familial likelihood for autism who do and do not receive later diagnoses. C_LIO_LIParent-reported and measure-defined regression in this high-likelihood sample have low concordance indicating that these may be different events in language development. C_LIO_LIThe presence of language regression was not associated with significantly higher odds of receiving an autism diagnosis. C_LIO_LIChildren who exhibit language regression continue growing and developing language and those with parent-reported regression display comparable language skills to children without language regression at three years of age. C_LI

Despite evidence for prenatal onset of abnormal head growth in children with autism spectrum disorder (ASD), fetal ultrasound studies in ASD are limited and controversial. We conducted a longitudinal matched case-sibling-control study on fetal ultrasound biometric measures from 174 ASD children, their own typically developed siblings (TDS; n=178) and other population-based typically developed children (TDP; n=176). During second trimester, ASD and TDS fetuses had significantly smaller biparietal diameter (BPD) than TDP fetuses (aORzBPD=0.685, 95%CI=0.527-0.890 and aORzBPD=0.587, 95%CI=0.459-0.751, respectively). Interestingly, sex had a significant effect on head growth with males having larger heads than females within and across groups. Also, males and females with ASD showed different head shapes which were inversely correlated with ASD severity across different gestation periods. Our findings suggest that abnormal fetal head growth is a familial trait of ASD, which is modulated by sex and is associated with the severity of the disorder.

ObjectivesThe male preponderance in autism spectrum conditions (ASC) prevalence is among the most pronounced sex ratios across different neurodevelopmental conditions. Here, we aimed to elucidate the relationship between autism and typical sex-differential neuroanatomy, cognition, and related gene expression. MethodsUsing a novel deep learning framework trained to predict biological sex, we compared sex prediction model performance across neurotypical and autistic males and females. Multiple large-scale datasets were employed at different stages of the analysis pipeline: a) Pre-training: the UK Biobank sample (>10.000 individuals); b) Transfer learning and validation: the ABIDE datasets (1,412 individuals, 5-56 years of age); c) Test and discovery: the EU-AIMS/AIMS-2-TRIALS LEAP dataset (681 individuals, 6-30 years of age) and d) Specificity: the Neuroimage and ADHD200 datasets (887 individuals, 7-26 years of age). ResultsAcross both ABIDE and LEAP we showed that features positively predictive of neurotypical males were on average more predictive of autistic males (P=1.1e-23). Features positively predictive of neurotypical females were on average less predictive of autistic females (P=1.2e-22). These accuracy differences in autism were not observed in individuals with ADHD. In autistic females the male-shifted neurophenotype was further associated with poorer social sensitivity and emotional face processing while also with associated gene expression patterns of midgestational cell types. ConclusionsOur results demonstrate a shift in both autistic male and female individuals neuroanatomy towards male-characteristic patterns associated with typically sex-differential, social cognitive features and related gene expression patterns. Findings hold promise for future research aimed at refining the quest for biological mechanisms underpinning the etiology of autism.

ObjectiveMaternal immune activation during pregnancy has been proposed as a mechanism linking prenatal inflammatory exposures to autism pathogenesis. While preclinical and epidemiological studies suggest a role for maternal inflammation and infection, findings in population-based cohorts are inconsistent. This study examined the associations between multiple prenatal inflammatory exposures and autistic traits, accounting for gene-environment interactions in the general pediatric population. MethodsWe leveraged data from 5,075 mother-child dyads participating in Generation R, a population-based pregnancy cohort in the Netherlands. Prenatal inflammatory exposures included 1) maternal serum cytokines; 2) high-sensitivity CRP; 3) self-reported fever during pregnancy; 4) a maternal polygenic score for CRP; and 5) a methylation profile score of CRP in cord blood. Child autistic traits were measured with the Social Responsiveness Scale at mean ages 6 and 13 years. Linear mixed models were applied to estimate associations adjusted for maternal, child and technical covariates. Interaction terms tested whether child polygenic score for autism moderated associations. ResultsNo significant associations were observed between prenatal inflammatory exposures and autistic traits, both as a continuous measure and above a clinical threshold. No evidence was found for interactions between prenatal inflammatory exposures and the child polygenic score for autism in influencing autistic traits. ConclusionOur findings suggest that typical fluctuations in maternal inflammation are unlikely to represent a major pathway linking prenatal environment to autism risk. We found no evidence that gene-environment interactions conferred additional risk for autistic traits.

Boys with fragile X syndrome (FXS), the leading known genetic cause of autism spectrum disorder (ASD), demonstrate significant impairments in social gaze and associated weaknesses in communication, social interaction, and other areas of adaptive functioning. Little is known, however, concerning the impact of behavioral treatments for these behaviors on functional brain connectivity in this population. As part of a larger study, boys with FXS (mean age 13.23 +/- 2.31 years) and comparison boys with ASD (mean age 12.15 +/- 2.76 years) received resting-state magnetic resonance imaging scans prior to and following social gaze training administered by a trained behavior therapist in our laboratory. Network-agnostic connectome-based predictive modeling (CPM) of pre-treatment RSFC data revealed a set of positive (FXS > ASD) and negative (FXS < ASD) edges that differentiated the groups significantly and consistently across all folds of cross-validation. Following administration of the brief training, the FXS and ASD groups demonstrated normalization of connectivity differences. The divergence in the spatial pattern of normalization response, based on functional connectivity differences pre-treatment, suggests a unique pattern of response to treatment in the FXS and ASD groups. These results support using connectome-based predictive modeling as an outcome measure in clinical trials.

Atypical eye gaze in joint attention is a clinical characteristic of autism spectrum disorder (ASD). Despite this documented symptom, neural processing of joint attention tasks in real-life social interactions is not understood. To address this knowledge gap, functional-near infrared spectroscopy (fNIRS) and eye-tracking data were acquired simultaneously as ASD and typically developed (TD) individuals engaged in a gaze-directed joint attention task with a live human and robot partner. We test the hypothesis that face processing deficits in ASD are greater for interactive faces than for simulated (robot) faces. Consistent with prior findings, neural responses during human gaze cueing modulated by face visual dwell time resulted in increased activity of ventral frontal regions in ASD and dorsal parietal systems in TD participants. Hypoactivity of the right dorsal parietal area during live human gaze cueing was correlated with autism spectrum symptom severity: Brief Observations of Symptoms of Autism (BOSA) scores (r = -0.86). Contrarily, neural activity in response to robot gaze cueing modulated by visual acquisition factors activated dorsal parietal systems in ASD, and this neural activity was not related to autism symptom severity (r = 0.06). These results are consistent with the hypothesis that altered encoding of incoming facial information to the dorsal parietal cortex is specific to live human faces in ASD. These findings open new directions for understanding joint attention difficulties in ASD by providing a connection between superior parietal lobule activity and live interaction with human faces. Lay SummaryLittle is known about why it is so difficult for autistic individuals to make eye contact with other people. We find that in a live face-to-face viewing task with a robot, the brains of autistic participants were similar to typical participants but not when the partner was a live human. Findings suggest that difficulties in real-life social situations for autistic individuals may be specific to difficulties with live social interaction rather than general face gaze.

Autism Spectrum Disorder (ASD) is a large set of neurodevelopmental disorders of complex aetiology. A mix of genetic and environmental factors are likely to cause ASD. Genetic risk for autism comes from common genetic variation. Genomic imprinting refers to genes that have different expression patterns according to the parent of origin - being silenced when imprinted. Paternally active genes increase resource extraction from the mother and reduce resource burden on the father. Children with ASD show consistent overgrowth during their first 1-2 years of life. Recently, it has been shown that children with higher birth weight and length have an increased risk of developing ASD. This overgrowth and apparent larger birth weight and length are consistent with the notion that a paternally biased genome might underlie the risk for ASD. The study compared height, weight, head circumference and thoracic circumference for age-matched (ages 4-8 years old) male children with ASD (n=30) with neurotypical children (n=33). No clinically significant differences were found among the two groups. After weaning, relative paternal contribution to a childs somatic development would increase, thus one would expect paternally active genes to start changing the childs behaviour, so as to make the child less demanding of resources (overall, and thus also on the father), with a counterweight represented by maternally active genes. A relative overabundance of paternally active genes would explain the data presented here, that shows children with ASD being no different from controls. Given the fact presented by other studies, that children with ASD seem to get a head start in growth, the lack of differences found in this 4-8 years old group indicates that children with ASD might actually fall behind in somatic growth, or at least stagnate by middle childhood.

Despite two decades of intensive investigation, the neurobiological underpinnings of autism spectrum disorder (ASD) defining symptoms remain elusive in large part because ASD represents a highly heterogeneous set of phenotypes, with wide-ranging genetic, neural and behavioural profiles that likely arise from multiple, partially overlapping mechanisms. Among various proposed mechanisms, the "excitation/inhibition (E/I) imbalance hypothesis" has been extensively researched, with substantial although sometimes conflicting evidence. Animal studies show that localised alterations in cortical E/I balance can disrupt network dynamics and reduce long-range connectivity, potentially explaining the weak central coherence that may be observed in ASD. However, neither the E/I imbalance hypothesis nor the weak central coherence hypothesis has provided a fully coherent explanation of ASDs neurobiology or proved useful in identifying reliable biomarkers. Here, we present a novel integrated hypothesis that unifies these theories, proposing that it is not the global average E/I ratio but rather its spatial heterogeneity across brain regions that may be altered in ASD and that disrupts the integration of brain network activity. To test this hypothesis, in a series of pre-registered analyses, we assessed EEG-derived E/I spatial autocorrelation in children with ASD (n = 248) and in neurotypical controls (n = 105). Our findings indicate significantly lower spatial autocorrelation in children with ASD, suggesting increased E/I variability across regions (estimate: -0.51, 95% CI: -0.69 - -0.33), both in wakefulness (total wakefulness EEGs n = 234; estimate -0.36, 95% CI: -0.63 - -0.081) and sleep EEG (total sleep EEGs n = 304; estimate -0.49, 95% CI: -0.75 - -0.23). Importantly, we validated these results in a large external dataset (ASD n = 230, NT n = 167; estimate: -0.22, 95% CI -0.33 - -0.02) compiled from the National Institute of Mental Health Data Archive (NDA), reinforcing the robustness of our findings. Moreover, when exploring diagnostic accuracy, EEG-based E/I spatial autocorrelation along with sex differentiated ASD from neurotypical status with fair performance (ROC AUC: 0.725, sensitivity: 0.89, specificity: 0.67), highlighting its potential for optimization as a cost-effective screening tool. Overall, these results illuminate a possible neurobiological signature of ASD and suggest that EEG measures of E/I spatial heterogeneity may serve as a viable biomarker for the condition.

BackgroundThe expression of autism traits sufficient to meet criteria for a diagnosis can occur early (by 3 years) or later (from mid-childhood onwards). It remains unknown whether variation in age of onset is due to clinical recognition or whether it reflects distinct biological pathways. One way of addressing this question is by investigating biological differences very early in development associated with age of onset. We use a prospective design to look at event related potentials to faces, one of the most robust biomarkers in autism. MethodsA sample of 102 infants (aged 6-10 months, 54% female) with an older autistic sibling had EEG recorded whilst viewing faces (faces versus noise; gaze towards versus away). Autism diagnostic assessments were conducted at three years and again in mid-childhood (aged 6-12 years), resulting in early diagnosed (at age 3; N=22), later diagnosed (at mid-childhood; N=21) and no autism (N=59) groups. ResultsWhile a short latency response (P1) does not associate with autism outcome, a mid-latency component (N290) associates with early onset autism only, and a later latency component (P400) associates with both early and later onset autism. ConclusionTemporal stages of face processing in infancy differentially associate with age of autism onset such that an earlier age of diagnosis is associated with earlier stage deviation within the event-related waveform. Early and later onset autism may represent different biological subtypes, with different early brain development, challenging the view of one etiological pathway and that variation in diagnostic age is solely due to clinical ascertainment.

Clinical research in neurodevelopmental disorders remains reliant upon clinician and caregiver measures. Limitations of these approaches indicate a need for objective, quantitative, and reliable biomarkers to advance clinical research. Extant research suggests the potential utility of multiple candidate biomarkers; however, effective application of these markers in trials requires additional understanding of replicability, individual differences, and intra-individual stability over time. The Autism Biomarkers Consortium for Clinical Trials (ABC-CT) is a multi-site study designed to investigate a battery of electrophysiological (EEG) and eye-tracking (ET) indices as candidate biomarkers for autism spectrum disorder (ASD). The study complements published biomarker research through: inclusion of large, deeply phenotyped cohorts of children with autism spectrum disorder (ASD) and typical development; a longitudinal design; a focus on well-evidenced candidate biomarkers harmonized with an independent sample; high levels of clinical, regulatory, technical, and statistical rigor; adoption of a governance structure incorporating diverse expertise in the ASD biomarker discovery and qualification process; prioritization of open science, including creation of a repository containing biomarker, clinical, and genetic data; and use of economical and scalable technologies that are applicable in developmental populations and those with special needs. The ABC-CT approach has yielded encouraging results, with one measure accepted into the FDAs Biomarker Qualification Program to date. Through these advances, the ABC-CT and other biomarker studies in progress hold promise to deliver novel tools to improve clinical trials research in ASD.

People with autism spectrum disorder (ASD) frequently exhibit challenging behaviors throughout the lifespan, which can have pervasive effects on quality of life. Challenging behaviors have been shown to change over time as a function of various individual-level factors (e.g., cognitive ability), yet research is primarily limited to parent-reported measures. To expand upon this work, the present study aimed to examine trajectories of teacher- and parent-reported challenging behaviors (i.e., hyperactivity, irritability, social withdrawal) and to test whether predictors including ASD features, verbal intelligence quotient, and consistency in reporter impact these trajectories among individuals with ASD or non-spectrum delays from ages 9 to 18. Multilevel models revealed that, according to both teacher and parent report, participants showed the greatest improvement in hyperactivity, less but still notable improvement in irritability, and stable levels of social withdrawal over time. Higher cognitive ability and fewer ASD features emerged as important individual differences related to fewer challenging behaviors. The multi-informant perspective and longitudinal design provide novel insight into the manifestations of these challenging behaviors across different contexts and across time. Findings highlight the importance of addressing challenging behaviors as these behaviors tend to persist throughout development in both home and school contexts, especially for children with particular diagnostic and cognitive profiles. Lay SummaryAccording to both teacher and parent report, youth with autism showed the greatest improvement in hyperactivity, less but still notable improvement in irritability, and stable levels of social withdrawal from school-age to adolescence. Fewer autism features and greater cognitive ability were related to fewer challenging behaviors. This studys use of multiple reporters (e.g., teachers and parents) across time provided insight into the persistence of challenging behaviors in the home and school settings and across development.

ImportanceMotor skill impairments affect up to 87% of children with autism spectrum disorder (ASD) and are associated with greater severity of repetitive behaviors. Yet, most research examining this relationship has treated ASD as a unitary condition. Understanding whether motor-behavior relationships differ by genetic etiology could inform stratified approaches to ASD research and clinical care. ObjectiveTo determine whether the relationship between motor function and restricted and repetitive behaviors (RRBs) differs between children with monogenic forms of ASD (SHANK3, DYRK1A, or SCN2A variants) and children with idiopathic ASD. Design, Setting, and ParticipantsMatched cohort cross-sectional study using data from the Simons Foundation Powering Autism Research for Knowledge (SPARK) database. Children with loss-of-function variants in SHANK3, DYRK1A, or SCN2A were matched to children with idiopathic autism and intellectual disability. Main Outcomes and MeasuresMotor function was assessed using the Developmental Coordination Disorder Questionnaire (DCDQ). Repetitive behaviors were assessed using the Repetitive Behavior Scale-Revised (RBS-R), with subscales categorized as lower-order (stereotyped, self-injurious) or higher-order (compulsive, ritualistic, sameness, restricted interests). The primary analysis compared motor-RRB correlations between groups. ResultsThe sample included 93 children with monogenic autism (SHANK3, n=34; DYRK1A, n=46; SCN2A, n=13) and 787 matched children with idiopathic ASD. In idiopathic ASD, motor function was negatively correlated with RRBs (r = -0.156); in monogenic ASD, this reversed to a positive correlation (r = +0.185; {Delta}r = 0.341, P = 0.002). This reversal was specific to higher-order RRBs (idiopathic r=-0.106; monogenic r=+0.234; {Delta}r=0.339, 95% CI 0.124-0.535, P=0.002) and was not observed for lower-order RRBs ({Delta}r=0.212, P=0.05). All three genes showed positive correlations (SHANK3 r=+0.033; DYRK1A r=+0.262; SCN2A r=+0.623) with no significant heterogeneity (P=0.153). Conclusions and RelevanceThe relationship between motor function and repetitive behaviors differs by genetic etiology, with children with monogenic ASD showing a positive motor-RRB correlation specific to higher-order behaviors, opposite to the negative correlation observed in idiopathic ASD. This reversal was consistent across three molecularly distinct genes. These findings support stratifying autism research and clinical care by genetic etiology. KEY POINTSO_ST_ABSQuestionC_ST_ABSDoes the relationship between motor function and restricted and repetitive behaviors (RRBs) differ between children with autism spectrum disorder (ASD) attributable to SHANK3, DYRK1A, or SCN2A variants and children with idiopathic ASD? FindingsWe conducted a matched cohort cross-sectional study comparing correlations between motor function and RRBs in children with monogenic ASD versus children with idiopathic ASD and intellectual disability. Motor function was negatively correlated with RRBs in children with idiopathic ASD but positively correlated in children with monogenic ASD. MeaningGenetic variants may alter behavioral organization, supporting the value of stratifying populations of individuals with ASD by genetic etiology in both research and clinical care.

Using data from a well-characterized longitudinal cohort, this study examined trajectories of restricted and repetitive behaviors (RRBs), specifically repetitive sensorimotor (RSM), insistence on sameness (IS), and verbal RRBs, as measured by the Autism Diagnostic Interview, Revised (ADI-R) from ages 2 to 19. Additionally, this study investigated relationships between RSM and IS trajectories and parent- and self-report depression and anxiety symptoms in early adulthood. Group-based trajectory modeling and multilevel modeling were used to investigate change in RRB subtypes. RSM and IS behaviors generally decreased from 2 to 19, though some participants experienced increases in these RRB subtypes from ages 2 to 9. 65% of this sample had sufficient verbal ability prior to age 19 to calculate trajectories of verbal RRBs. Of this subset, 49% had few to no verbal RRBs throughout development; in contrast, the remaining 51% experienced increasing verbal RRBs from 2 to 9, followed by a plateau in verbal RRBs from 9-19. Higher ADOS Social Affect (SA) CSS scores in early childhood were linked with more RSM symptoms across development, but not related to the IS and Verbal trajectories. Non-verbal IQ from early childhood was not connected to change in any of the identified RRB trajectories. There were no associations between IS trajectories and internalizing symptom in early adulthood. However, preliminary data suggests that a Moderate-Decreasing pattern of RSM development may be linked to anxiety in early adulthood. These findings illustrate continuity and change in a core ASD symptom domain, RRBs, from early childhood to early adulthood.

BackgroundAlterations in the brains oxytocinergic system have been suggested to play an important role in the pathophysiology of autism spectrum disorder (ASD), but insights from pediatric populations are sparse. MethodsWe examined salivary oxytocin in school-aged children with (n=80) and without (n=40) ASD (boys/girls 4/1), as well as characterizations of DNA methylation (DNAm) of the oxytocin receptor gene (OXTR). Cortisol levels were also assessed to examine links between the oxytocinergic system and hypothalamic-pituitary-adrenal (HPA) axis signaling. ResultsChildren with ASD displayed altered (diminished) oxytocin levels in the morning, but not in the afternoon, after a mildly stress-inducing social interaction session. Notably, in the control group, higher oxytocin levels were predictive of lower stress-induced cortisol, likely reflective of a protective stress-regulatory mechanism for buffering HPA stress activity. In children with ASD, on the other hand, a more reactive stress regulatory mechanism was evident, involving a significant rise in oxytocin levels from the morning to the afternoon upon stress-induced cortisol release, i.e., to reactively cope with heightened HPA activity. Regarding epigenetic modifications, no overall pattern of OXTR hypo- or hypermethylation was evident in ASD. In control children, a notable association between OXTR methylation and levels of cortisol was evident, likely indicative of a compensatory downregulation of OXTR methylation (higher oxytocin receptor expression) in children with heightened HPA axis activity. ConclusionTogether, these observations bear important insights into altered oxytocinergic signaling in ASD, which may aid in establishing relevant biomarkers for diagnostic and/or treatment evaluation purposes targeting the oxytocinergic system in ASD.

Longitudinal, prospective analyses of marital status in parents of individuals with autism are needed. We describe the timing of divorce, and the factors that contribute to divorce in a longitudinal sample of families of individuals with autism. Participants included parents of 219 children, initially referred for autism and other developmental delays, followed to age 30 years. Approximately 36% of individuals with autism in our sample experienced a parental divorce by age 30. Higher rates of divorce were associated with maternal education, race and age at childs birth, as well as autism symptom severity and diagnosis. Divorces were most common in early years (under age 5) and also in the teenage years and beyond (over age 15). After age 15, higher risk was associated with higher cognitive ability and daily living skills, and being a multiplex family. Results suggest that divorce risk in families of children with autism remains high through childhood into early adulthood. Understanding factors related to changes in marital status may help us better support families across time.

Infants born preterm are at a significantly higher likelihood of having autism spectrum disorder (ASD). Preterm birth and ASD are both associated with neurological differences, notably autonomic nervous system (ANS) dysfunction, pointing to preterm ANS dysfunction as a potential pathway to ASD, particularly in VPT infants. In this study, a subset of very preterm (VPT) infants enrolled in a large, multisite clinical trial were enrolled in this study at birth (N=20). Continuous measures of minute-by-minute thermal gradients, defined by the difference between central and peripheral temperatures, and hour-by-hour abnormal heart rate characteristics (HRCs) were collected from birth-28 days (>40,000 samples/infant). Following NICU discharge, standardized measures of cognition, language, and motor skills were collected at adjusted ages 6, 9, and 12 months. At 12 months, assessments of social communication and early ASD symptoms were administered. Results suggest significant ASD concerns for half of the sample by 12 months of age. Neonatal abnormal HRCs were strongly associated with 12-month ASD symptoms (r=0.81, p<.01), as was birth gestational age (GA), birth weight (BW), and abnormal negative thermal gradients. ANS measures collected in the first month of neonatal life, more than a year prior to the ASD evaluation, were surprisingly strong predictors of ASD. This study highlights complementary ANS measures that describe how ANS dysfunction, likely resulting from an imbalance between the parasympathetic and sympathetic systems, may impact very early regulatory processes for neonates who later develop ASD. This finding offers a promising avenue for researching ANS-related etiological mechanisms and biomarkers of ASD.

Many autistic people have challenges with adaptive function, impacting education, employment and independent-living goals. Adaptive function outcomes of autistic people vary considerably, which makes planning for future needs challenging. Here, using a developmentally sensitive approach, we investigated if cortico-cortical functional connectivity - a core neurobiological feature that differs in autism - could predict longitudinal changes in adaptive function in autistic people. Using electroencephalography in 150 autistic and 159 non- autistic participants aged 6-31 years, we investigated if mean degree and network organisation (small-world index) predict longitudinal changes in adaptive function over 19-months. We assessed both metrics for properties desired in prognostic biomarkers: reliability and convergence with biology (polygenic variation). We found that small-world index significantly predicted changes in adaptive function in autistic people across the entire age-range. Predictive performance was best in 15-21-year-olds, where small-world index and mean degree explained 30-33% of additional variance in outcomes, outperforming measures of intelligence and autistic features. In categorising binary (improved versus not-improved) outcomes, the model containing mean degree had an AUC of 0.80 [95% CI: 0.63-0.97] in 15-21-year-olds, while that containing small-world index had an AUC of 0.76 [95% CI: 0.63-0.89] across the 6-31- year age-range. Both metrics demonstrated high test-retest reliability and significant associations with polygenic variation in brain volume. We demonstrate the first evidence that electroencephalography-derived functional connectivity metrics show promise as prognostic biomarkers of adaptive function in autistic people. Potential precision-medicine applications include stratifying participants in clinical trials and identifying those at risk of declining function in clinical settings.

Altered motivational drives and aberrant reward system function may contribute to the social impairments observed in autism spectrum disorders (ASD). Pupil metrics have been highlighted as peripheral indicators of autonomic arousal and reward system function, specifically noradrenergic and dopaminergic activity that influence motivational drive states. However, research on individual differences in the neurobiological correlates of reward responsivity and clinically relevant features associated with ASD is sparse. The goal of the current study was to examine the relationship between measures of sensitivity to punishment and reward, ASD features, and resting as well as functional pupil response metrics across a clinically heterogeneous pediatric sample. We assessed whether quantitative features of reward sensitivity are linearly related to core clinical features of ASD. Pupil metrics were measured using a passive eye tracking task. Scores on a parent-report measure of punishment and reward sensitivity were found to be positively correlated with ASD features. Given these relationships, we assessed whether pupil measurements could be used as a neurobiological correlate of reward sensitivity and predictor of clinically significant ASD traits. In a logistic regression model, we find that the amplitude of pupil dilation, along with sex and full-scale IQ, could be used to correctly classify 84.9% of participants as having an ASD diagnosis versus not having an ASD diagnosis. This research highlights individual differences of reward sensitivity that scale with ASD features. Furthermore, reported results emphasize that functional pupil response metrics and other objective patient-level variables can be used together as predictors of ASD diagnostic status.

Autism is a highly heterogenous neurodevelopmental condition that, in some individuals, can significantly impair quality of life. Early identification followed by timely intervention is crucial for enhancing cognitive outcomes in children with developmental delay. Yet, marked variability in intervention response remains insufficiently understood. Characterizing this heterogeneity is essential for informing intervention strategies tailored to each childs developmental profile. Here, we aim to identify subgroups based on intervention response trajectories and investigate subgroup-specific predictors of intervention outcomes. We analyzed longitudinal data from 107 children (1.7-3.5 y.o.) receiving 1.5-2 years of Early Start Denver Model (ESDM) intervention. Children gained an average of 19 points in Developmental Quotient (DQ), significantly improved their adaptive skills and showed less autism features. Using latent class regression model, we identified 3 classes: Progressive Group A (35.5% of participants), which had the highest baseline cognitive scores (mean DQ: 78) and gained 27 DQ points in average throughout the intervention; Progressive Group B (34.6% of participants), which showed significant developmental delay at baseline (mean DQ: 62) and gained 25 DQ points; and the Stable Group (29.9% of participants), which also started with significant developmental delay (mean DQ: 47) and showed steady and modest improvement in cognitive scores throughout the intervention. Among the subgroup-specific predictors of better cognitive outcome were younger age for the Progressive Groups and fewer restrictive and repetitive behaviors for the Stable Group. Our results support previous findings of ESDMs efficacy, replicated in a larger sample of 107 children and reveal the association of subgroup-specific baseline factors with cognitive evolution, offering a first step towards personalizing interventions. Lay SummaryO_LIIn a group of 107 autistic children receiving Early Start Denver Model intervention we observed significant gain in cognitive and adaptive skills as well as reduced level of autism features. C_LIO_LIWe identified three subgroups of response with different developmental trajectories. Different factors at baseline were selected as predictors of the cognitive outcome for each subgroup. For two of the subgroups, younger age at the start of intervention was linked to better cognitive outcomes. C_LI