Face perception is essential for social interaction, but its neural basis and development in autism remain unclear. Using multivariate pattern analysis on high-density electroencephalography data from a large pediatric case-control observational study (N = 399), we found that autistic children exhibited less distinct face- and identity-selective neural representations than neurotypical (NT) children. During specific stages of face processing, the neural representations for faces, inverted faces and houses were significantly more distinct in NT compared with autistic children. Across developmental age groups (6-7-, 8-9- and 10-11-year-olds), NT children showed increasing neural specialization across age such that face- and identity-selective neural representations became more distinct across age, a pattern not observed in autistic children. These findings provide insight into the neural processes underlying face perception differences in autism and underscore the importance of understanding developmental trajectories of face-selective neural representations.

Tourette syndrome (TS) is heterogeneous and frequently co-occurs with attention-deficit/hyperactivity disorder (ADHD) and other recorded neurodevelopmental diagnoses. We examined whether early parent-report questionnaire profiles in the Norwegian Mother, Father and Child Cohort Study (MoBa) differed across later TS diagnostic subgroups, focusing on developmental phenotyping rather than individual-level prediction. The analytic sample included 92,875 children with linked registry follow-up and at least one retained questionnaire predictor at 6, 18, or 36 months. For this project, registry access was limited to ICD-10 F70-F99 diagnoses; the reference group therefore comprised children with no recorded F70-F99 diagnosis in the available project extract, not children known to be free of all psychiatric or somatic diagnoses. Age-specific principal component analysis (PCA), item response theory (IRT), and raw-item profiles were used to describe early questionnaire dimensions relative to the no-recorded-F70-F99 reference group. Twenty-one PCA domains were retained. Differences were small at 6 months and most interpretable at 36 months. TS with ADHD showed the clearest preschool regulation, attention, and externalizing profile, whereas TS with other recorded F70-F99 diagnoses showed broader language, motor, social-communication, sociability/activity, and developmental-concern elevations. TS without other recorded F70-F99 diagnoses showed smaller, more focal elevations. Secondary separability checks were modest, supporting group-level profile differences rather than screening-level prediction.

Transient beta events (TBE) during electroencephalography (EEG) reflect thalamocortical activity, bridging genotype to phenotype and impacting sensory responsivity. Compared to typically developing controls, we found elevated TBE rate in some children with idiopathic Autism Spectrum Disorder (ASD) and a majority of children with Phelan-McDermid Syndrome, Rett Syndrome, and SYNGAP1-related disorder. TBE rate thus offers promise as a stratification biomarker with divergent and convergent properties across ASD and neurogenetic conditions, respectively.

Functional near-infrared spectroscopy (fNIRS) is a promising modality for autism spectrum disorder (ASD) classification, yet existing approaches assume temporally aligned evaluation. In practice, the optimal observation window varies across subjects due to differences in hemodynamic delay and neurovascular coupling, creating a temporal distribution shift that degrades performance. We formalize this as a cross-time-window transfer problem, introducing a protocol that varies window length (2.5–10 s) and offset within biological motion trials. Using topographic map representations of fNIRS recordings, we benchmark three vision architectures under two zero-shot baselines and eight adaptation strategies under leave-one-subject-out cross-validation (N=124). Key findings: (1) zero-shot cross-window accuracy is near chance (54–69%); (2) ≈5% subject-specific fine-tuning recovers 90–96%, while a subject-specific upper bound reaches 97–100%, identifying inter-subject variability as the dominant barrier; (3) domain-adversarial and self-supervised strategies achieve 78–90% without target-subject data; and (4) discriminative information is recoverable from windows as short as 2.5 s. These findings provide a practical roadmap for deploying fNIRS-based ASD classifiers under realistic temporal variability.

This review provides an overview of eye-tracking tasks used to investigate joint attention (JA) in autistic children by examining key methodological dimensions, including the structure of the JA probe, design of the JA task, and gaze metrics derived to quantify JA. A systematic search was conducted following the PRISMA guidelines. A total of 26 studies were included in this review. The findings highlight a heavy focus on responding to joint attention (RJA), with limited emphasis on initiating joint attention (IJA). There was a clear preference for dynamic yet non-interactive social stimuli without environmental context, accompanied by variability in deictic cues for RJA and prompts for IJA tasks. Nine gaze metrics were identified, showing variability in their computation and prerequisites. These findings have implications for the conceptualization, design, and refinement of JA eye-tracking tasks and their applications for autistic children.

Autism is characterized by marked heterogeneity in behavioral presentation and high rates of co-occurring psychiatric symptoms, which hinder diagnostic precision, personalized intervention, and long-term quality of life. Approach and withdrawal behaviors—subserved by core motivational systems underlying action and emotion—may serve as transdiagnostic processes linking autism with common co-occurring conditions in childhood. Guided by this framework, we examined how autism-related approach–withdrawal behaviors interrelate and connect to internalizing and externalizing symptoms. Using data from 280 autistic children aged 6 to 11 years enrolled in the Autism Biomarkers Consortium for Clinical Trials, we constructed a Gaussian graphical model of approach–withdrawal behaviors. Core behaviors were identified using expected influence centrality. Autism, when conceptualized as a system of interconnected approach–withdrawal behaviors, was positively associated with common co-occurring psychiatric conditions, with strongest associations observed for anxiety and attention-deficit/hyperactivity disorder. Affect regulation–related nodes were most relevant to internalizing symptoms, whereas arousal regulation and sensory nodes were uniquely related to externalizing symptoms. These findings integrate transdiagnostic theories of approach-withdrawal motivation with network analysis and highlight clinically relevant targets for diagnostic refinement and intervention.

Autism spectrum disorder is a heritable neurodevelopmental condition affecting approximately 3% of children that presents with core behavioral features and a range of possible comorbidities, including intellectual disability. While common variants contribute substantially to autism liability, the discovery of specific autism-associated genes has largely been driven by studies of rare and de novo variants. Many of these genes are also linked with broadly defined developmental disorders, but their involvement in other conditions has not been mapped at scale. Here, we analyze autosomal rare coding variation from 62,429 individuals with autism from research and clinical cohorts to identify 253 autism-associated genes at an estimated false discovery rate < 0.001. We cluster them based on association evidence from large-scale studies of developmental disorders, schizophrenia, bipolar disorder, and epilepsy, generating six clusters of genes with differing biological pathway enrichments and patterns of comorbidities. Investigating rare variant associations in the population using the UK Biobank and All of Us, we identify autism-associated genes displaying pleiotropy across physiological systems. In addition, we report 497 genes impacting development in a meta-analysis with 26,109 published developmental disorders samples. Collectively drawing upon data from over 1.5 million individuals, our study finds that rare variants across hundreds of genes contribute to autism with variable phenotypic outcomes.

Eye tracking has emerged as a powerful tool for advancing autism research, including diagnostics and interventions. However, studies on joint attention (JA) in autistic children have predominantly concentrated on the responding to joint attention (RJA) construct, with limited focus on the initiating joint attention (IJA) construct. Moreover, despite the interactive nature of JA, researchers have often relied on passive paradigms to study JA in this population. To address these gaps, we developed a gaze-contingent eye-tracking battery targeting three developmentally appropriate JA skills for young children: RJA, IJA to request, and IJA to comment or reference. The development process was multifaceted and iterative, involving a series of collaborative steps and the allocation of various resources. These steps included determining the motion format and stimulus type, designing and prototyping the stimuli, recruiting an actor to serve as a communication partner in the stimuli, recording and editing videos for the stimuli, and building and test-running the battery. We developed the Interactive Eye Tracking for Joint Attention (IET-JA) battery, which consists of 32 JA stimuli: 16 RJA, 8 IJA-Request, and 8 IJA-Comment/Reference. The stimuli are dynamic (i.e., videos) and feature a preprogrammed interactive human communication partner who is responsive to the participant’s gaze. The IET-JA takes approximately 8 minutes to complete, and its duration is expected to vary based on the participant’s level of engagement. Implications for advancing methodologies, fostering team science, and enhancing iterative processes are discussed.

Reduced social attention (SA) is a hallmark feature of autism that is foundational to social communication and interaction. However, emerging evidence suggests that reduced SA may not be uniformly expressed across the autism spectrum. Sex and cognitive ability (IQ) have been identified as relevant stratification variables, potentially moderating SA in autistic children. We examined differential patterns of SA stratified by sex and cognitive ability in a large, longitudinal sample of autistic and neurotypical (NT) children (ages 6-11 years) from the Autism Biomarkers Consortium for Clinical Trials (ABC-CT), examining SA across four measurement timepoints (Baseline, 6 weeks, 6 months, and 4 years). SA was measured using the oculomotor index of gaze to human faces (OMI) derived from standardized eye-tracking (ET) assays, with children stratified by sex (male, female) and IQ (IQ ≥ 85, IQ < 85). Linear mixed-effects models were used to evaluate the effects of sex, IQ, and timepoint on SA, with additional analyses assessing the stability and clinical associations of OMI. Autistic females did not exhibit lower OMI scores compared to autistic males; however, relative to sex-matched NT children, autistic females showed approximately twice the reduction in OMI observed in autistic males. Likewise, the reduction in OMI associated with autism (i.e., ASD < NT) was larger in autistic children with below-average IQ than in those with average or above-average IQ. These differences were relatively stable across measurement timepoints, and OMI was more strongly associated with clinical features like face memory and adaptive behavior in autistic males and the average or above-average IQ autism subgroup. While reduced SA is a general feature of autism, its expression is stratified by sex and cognitive ability. These findings highlight the importance of subgroup stratification to better understand heterogeneity in autism and improve the utility of SA as a therapeutic biomarker.

A growing body of work on robot-mediated therapy suggests that robots can elicit engagement and novel social behaviors among users with autism. In this narrative review, we trace the full historical arc of this field to date, from its inception in 2001 to 2024, covering 304 studies that present a robot for autism support. Early work largely consisted of short, highly structured sessions conducted in controlled laboratory or clinical environments. More recent research has shifted toward longer-term, real-world deployments in which robots operate with greater autonomy and engage users over multiple days or weeks. However, evidence for lasting and generalized benefits is still limited. The literature also remains focused primarily on children, with comparatively little research involving adults or individuals across a wider range of support needs. Despite rapid growth over these past two decades, the research remains fragmented across disciplinary boundaries, with robotics and clinical communities advancing similar goals without a cohesive, shared research framework. To address this, we offer a translational roadmap that details what interventions work, for whom, under which conditions, and why. We highlight current methodological gaps, outline key design considerations, and propose priorities for future research.

Social perception and attention markers have been identified that, on average, differentiate autistic from non-autistic children. However, little is known about how these markers predict behavior over time at both short and long time intervals. We conducted a large multisite, naturalistic study of 6- to 11-year-old children diagnosed with ASD (n = 214). We evaluated three markers of social processing: social perception via the ERP N170 Latency to Upright Faces; social attention via the Eye Tracking (ET) OMI (Oculomotor Index of Gaze to Human Faces) that captures percent looking to faces from three tasks; and social cognition via the NEPSY Face Memory task. Each was evaluated in predicting social ability and autistic social behaviors derived from parental interviews and questionnaires about child behavior at + 6 months (T3) and + 4 years (T4). Adjusting for baseline performance, time between measurements, age, and sex, our results suggest differential prognostic relations for each of the markers. The ERP N170 Latency to Upright Faces showed limited prognostic relations, with a significant relation to short term changes in face memory. The ET OMI was related to face memory over both short and long term. Both the ET OMI and Face Memory predicted long-term autistic social behavior scores. In the context of a large-scale, rigorous evaluation of candidate markers for use in future clinical trials, our primary markers had significant but small-effect prognostic capability. The ET OMI and Face Memory showed significant long-term predictive relations, with increased visual attention to faces and better face memory at baseline related to increased social approach and decreased autistic social behaviors 4 years later.

BACKGROUND:Social anhedonia, indicating reduced pleasure from social interaction, is heightened in autistic youth and associated with increased internalizing symptoms transdiagnostically. The stability of social anhedonia over time and its longitudinal impact on internalizing symptoms in autism have never been examined. METHODS:Participants were 276 autistic children (Mage = 8.60, SDage = 1.65; 211 male) with IQ ≥ 60 (MIQ = 96.74, SDIQ = 18.19). Autism severity was measured using the Autism Diagnostic Observation Schedule, Second Edition. Caregivers completed the Child and Adolescent Symptom Inventory, Fifth Edition (CASI-5) at baseline, 6 weeks, and 6 months. The CASI-5 includes a social anhedonia subscale derived from relevant items across domains. ICC (Intraclass Correlation Coefficient) analysis assessed stability, while cross-lagged panel models examined associations among social anhedonia, depression, and social anxiety across time. RESULTS:At baseline, social anhedonia correlated with autism severity, as well as parent-reported social anxiety and depression. Social anhedonia showed relative stability (ICC = 0.763) over 6 months, with a significant decline between baseline and 6 weeks (β = -0.52, p < .001). Cross-lagged models revealed a bidirectional relationship between social anhedonia and depression over time, while social anxiety displayed concurrent, but not predictive, associations across time. CONCLUSIONS:Social anhedonia demonstrated stability over 6 months, suggesting that it may be a relatively stable characteristic in autistic children. Concurrent relationships were observed between social anhedonia and depression, as well as social anxiety and attention-deficit/hyperactivity disorder. Only depression demonstrated a bidirectional longitudinal association with social anhedonia. This bidirectional relationship aligns with developmental models linking early negative social experiences to subsequent internalizing symptoms in autistic children, underscoring the clinical significance of social anhedonia assessment in this population.

Social attention, including shared attention and social orienting, is essential for positive social interactions. Although early visual social attention is often quantified using eye tracking, these indices may not consistently reflect cognitive engagement. Heart rate defined sustained attention (HRDSA) is a physiological measure that can index cognitive engagement alongside visual attention, leading to more comprehensive assessments of attentional processes that are particularly important in young, neurodiverse children with high support needs, including those with autism and fragile X syndrome (FXS). The present study examined visual and heart-defined measures of social attention to the Selective Social Attention task, a video-based assay of social attention, in children with autism, FXS, and neurotypical development. Linear mixed models examined group and condition effects in multiple cardiac indices and overall looking at the scene. Findings suggest that, overall, children across all groups engaged similarly across the experiment in most dimensions of HRDSA, and consistent with previous work, autistic children spent less time visually attending to the scene than either other group. HRDSA was positively associated with visual social attention. Combining physiological and visual attention measures may elucidate the complex nature of social attention and be especially valuable for neurodiverse children when typical assessments are inaccessible.

BACKGROUND: Reduced social attention-looking at faces-is one of the most common manifestations of social difficulty in autism that is central to social development. Although reduced social attention is well characterized in autism, qualitative differences in how social attention unfolds across time remains unknown. METHODS: We used a computational modeling (i.e., hidden Markov modeling) approach to assess and compare the spatiotemporal dynamics of social attention in a large, well-characterized sample of children with autism (n = 280) and neurotypical children (n = 119) (ages 6-11) who completed 3 social eye-tracking assays at 3 longitudinal time points (baseline, 6 weeks, 24 weeks). RESULTS: Our analysis supported the existence of 2 common eye movement patterns that emerged across 3 eyetracking assays. A focused pattern was characterized by small face regions of interest, which had high a probability of capturing fixations early in visual processing. In contrast, an exploratory pattern was characterized by larger face regions of interest, with a lower initial probability of fixation and more nonsocial regions of interest. In the context of social perception, children with autism showed significantly more exploratory eye movement patterns than neurotypical children across all social perception assays and all 3 longitudinal time points. Eye movement patterns were associated with clinical features of autism, including adaptive function, face recognition, and autism symptom severity. CONCLUSIONS: Decreased likelihood of precisely looking at faces early in social visual processing may be an important feature of autism that is associated with autism-related symptomology and may reflect less visual sensitivity to face information.

Dynamic eye-tracking paradigms are an engaging and increasingly used method to study social attention in autism. While prior research has focused primarily on younger populations, there is a need for developmentally appropriate tasks for older children. This study introduces a novel eye-tracking task designed to assess school-aged children’s attention to speakers involved in conversation. We focused on a primary outcome of attention to speakers’ faces during conversation between three actors and during emulated bids for dyadic engagement (dyadic bids). In a sample of 161 children (78 autistic, 83 neurotypical), children displayed significantly lower overall attention to faces compared to their neurotypical peers (p <.0001). Contrary to expectations, both groups demonstrated preserved attentional responses to dyadic bids, with no significant group differences. However, a divergence was observed following the dyadic bid: neurotypical children showed more attention to other conversational agents’ faces than autistic children (p =.017). Exploratory analyses in the autism group showed that reduced attention to faces was associated with greater autism features during most experimental conditions. These findings highlight key differences in how autistic and neurotypical children engage with social cues, particularly in dynamic and interactive contexts. The preserved response to dyadic bids in autism, alongside the absence of post-bid attentional shifts, suggests nuanced and context-dependent social attention mechanisms that should be considered in future research and intervention strategies.

Event-related spectral perturbations (ERSPs) capture dynamic changes in electroencephalography (EEG) power across frequency and trial time. Even though they are obtained at the trial level, they are commonly averaged across trials and analyzed at the subject level for enhancing the signal-to-noise ratio. While evoked activity is stimulus-locked, representing the brain’s predictable response to stimuli, induced signals that are not strictly locked to stimulus presentation are thought to be generated by higher-order processes, such as attention and integration. Motivated by joint modeling of multilevel (trials nested in subjects) and multivariate (evoked and induced) ERSP data from a visual-evoked potentials (VEP) task, we propose a multilevel multivariate functional principal components analysis (FPCA) for high-dimensional functional outcomes as a function of time and frequency. The proposed estimation procedure utilizes multilevel univariate FPCA decompositions along each variate of the multivariate outcome using fast covariance estimation and incorporates the dependency across outcome variates at each level of the data. Hence, the proposed approach for multilevel multivariate FPCA can efficiently scale up to higher-dimensional functional outcomes and increasing number of variates in the multivariate functional outcome vector. Extensive simulations show the efficacy of the proposed approach, while applications to VEP data lead to new insights on autism-specific neural activity patterns. The autistic group shows significantly lower evoked and higher induced gamma power compared to the neurotypical group. In addition, while subject level variation is dominated by variation in the stimulus-locked evoked signal in neurotypical development, it is dominated by induced power in autism.

Background:The Interactive Eye Tracking for Joint Attention (IET-JA), a child-focused battery consisting of video-format gaze-contingent eye-tracking tasks featuring a human communication partner, was developed to address gaps in eye-tracking research on JA in autistic children. Although JA research has often concentrated on younger populations due to its early developmental significance, studies indicate that JA difficulties associated with autism persist into adulthood, highlighting the value of examining these traits later in life. Thus, this study adapted the IET-JA for adolescents and adults (IET-JA-A) by incorporating additional attentional demands to introduce controlled attentional variations and explore whether these interactive eye-tracking measures of JA associate with autism traits in adulthood. Methods:A total of 81 young adults (M age = 19.32 years, range = 18-24 years), with a broad range of autism traits, completed the IET-JA-A. Generalized and linear mixed modeling were employed to address the aim. Results:The IET-JA-A measures of responding to joint attention and initiating joint attention to comment/reference (protodeclarative) were associated with autism traits, while the IET-JA-A measures of initiating joint attention to request (protoimperative) were not associated with autism traits. Conclusions:Findings advance our understanding of JA linked with autism traits in adulthood and showcase the feasibility of interactive eye-tracking methodologies in JA research. Limitations and implications for research are discussed.

Objective: Autism is characterized by differences in social functioning and many autistic individuals exhibit co-occurring features, such as aggression. While previous research has examined correlations between aggression and social skills cross sectionally, research is needed to better understand the interrelations between these behaviors over time. Method: In a sample of 280 autistic children (ages 6-11 years), parallel process growth modeling and regression were used to examine changes in social skills and aggression across three time points, as well as the interrelations between social skills and aggression. A confirmatory factor analysis (CFA) was also conducted on the measure of aggression (the Aberrant Behavior Checklist), which yielded separate factors for aggression towards others and self-injurious behaviors (SIB). Results: Over time, aggression towards others decreased and social skills improved, but SIB did not change. While initial levels of social skills, aggression towards others, and SIB were inter-related, there were no significant relations between initial scores and subsequent changes in any domain. Conclusions: Interventions for aggression in school-aged autistic children should consider aggression towards others, SIB, and social skills as separate targets, as changes in one domain did not appear to be related to changes in other domains.

Atypical gaze behavior is a diagnostic hallmark of Autism Spectrum Disorder (ASD), playing a substantial role in the social and communicative challenges that individuals with ASD face. This study explores the impacts of a month-long, in-home intervention designed to promote triadic interactions between a social robot, a child with ASD, and their caregiver. Our results indicate that the intervention successfully promoted appropriate gaze behavior, encouraging children with ASD to follow the robot's gaze, resulting in more frequent and prolonged instances of spontaneous eye contact and joint attention with their caregivers. Additionally, we observed specific timelines for behavioral variability and novelty effects among users. Furthermore, diagnostic measures for ASD emerged as strong predictors of gaze patterns for both caregivers and children. These results deepen our understanding of ASD gaze patterns and highlight the potential for clinical relevance of robot-assisted interventions.

Computer-vision-based eye tracking is poised to accelerate educational personalization through accessible quantification of learning and neurodevelopment. Recently, gaze patterns related to social communication were found to be more strongly associated in monozygotic (identical) versus dizygotic (fraternal) twins, suggesting genetic contributions. However, the intrinsic characteristic of blinking, which also reflects development and cognition, remains underexplored. This study uses computer-vision-based blink detection to characterize blinking in twins enrolled in a remote tablet-based study of infant and toddler attention. DBSCAN face-fingerprint clustering and caffemodel age prediction separated parent and child faces. Facial landmarks and head orientation were extracted using RetinaFace. Blink detection relied on fine-tuned EfficientNet-B4 refined by xgboost. Strong associations in blink probabilities were found across twins. Blink rates were correlated in monozygotic but not dizygotic twins. This work suggests that computer vision-derived blink measures may reflect genetic influences and expand a toolbox for attention-related quantification of intrinsic human variation.