1. Alzahrani AR, Alboaneen D, Alzahrani IR. GeneticNAS: a novel self-evolving neural architecture for advanced ASD screening. Sci Rep. 2026.

The early identification of Autism Spectrum Disorder (ASD) remains a critical challenge in neurodevelopmental research, with current diagnostic processes often delayed by subjective assessments and limited clinical resources. This paper presents a memory-efficient Neural Architecture Search framework that autonomously identifies optimum neural network structures for ASD classification. Unlike existing genetic algorithm-based NAS approaches requiring over 16GB GPU memory, our framework achieves 76% memory reduction while maintaining superior performance. Our approach presents three key innovations: (1) a novel search space integrating simple, residual, and bottleneck operations with [Formula: see text] complexity for L layers; (2) a memory-efficient genetic algorithm that decreases GPU memory consumption by [Formula: see text] relative to current methodologies while preserving search efficacy; and (3) an adaptive fitness function that equilibrates model performance with computational complexity. Through comprehensive experiments utilizing a substantial dataset ([Formula: see text]; [Formula: see text], [Formula: see text]), our methodology attained a classification accuracy of [Formula: see text] ([Formula: see text] CI: 94.89-[Formula: see text]) and area under the Receiver Operating Characteristic (ROC) curve of 0.986, which markedly surpassed existing state-of-the-art techniques (traditional CNN: 92.3%, ResNet-based: 94.1%, LSTM: 93.7%). The framework achieves this performance with 2.8M parameters and 15ms processing time per sample, demonstrating practical viability for clinical deployment in resource-constrained settings where current diagnostic procedures extend 4-5 years after symptom onset.

Lien vers le texte intégral (Open Access ou abonnement)

2. DiCriscio AS, Little JA, Troiani V. Ocular phenotypes associated with autism and atypical neurodevelopment: Insights from electronic health records. Vision Res. 2026; 242: 108767.

Atypical visual perception is often described in autism spectrum disorder (ASD); however, few studies have characterized ocular conditions in ASD using basic vision metrics such as those collected in routine eye exams. The current study uses electronic health record (EHR) codes to establish ocular phenotypes across individuals with and without neurodevelopmental diagnoses, including ASD. Using a population health approach, we assessed ocular conditions (identified based on medical codes from the EHR) in N = 7518 pediatric patients across 4 groups: n = 1196 with ASD, n = 156 with Intellectual Disability (ID), n = 347 with Language Disorder (LD), and n = 5819 matched controls (MC). We grouped and summarized ocular conditions across 5 ocular classes, including: (1) Visual impairment; (2) Refractive error, Accommodative & Vergence disorders; (3) Eye movements, Strabismus & Oculomotor Disorders; (4) Retinal disorders & Ocular disease; (5) Photosensitivity & Atypical Pupil response. We find an increased rate of ocular conditions in diagnostic groups compared to matched controls across classes 1 and 3. This study highlights the use of EHR data to curate ocular condition metrics collected in clinical care. The characterization of ocular anomalies across categories using EHR data offers a scalable method to improve our understanding of vision phenotypes that may be present in children with ASD and other neurodevelopmental differences.

Lien vers le texte intégral (Open Access ou abonnement)

3. Fan Y, Ai L, Tian Y. VaeTF-A community-aware perceptual architecture for detecting autism spectrum disorders using fMRI. Cogn Neurodyn. 2026; 20(1): 29.

Autism Spectrum Disorder (ASD) is a complex neurodevelopmental disorder, and the existing clinical diagnosis mainly relies on subjective behavioral assessment and lacks objective biomarkers. This paper proposes a hierarchical deep learning architecture, VaeTF, incorporating community-aware mechanisms based on resting-state functional magnetic resonance imaging (rs-fMRI) data. VaeTF introduces a priori knowledge of the functional community, extracts localized features through a variational auto-encoder (VAE), captures global dependencies across brain regions using the Transformer module, and incorporates an improved pooling mechanism to enhance the expressive power and model generalization performance. Experimental results on the ABIDE database show that VaeTF achieves 71.4% accuracy in ASD and typically performs well in group classification tasks. Further feature weighting analysis reveals that VaeTF is capable of identifying local functional abnormalities and cross-network functional synergistic dysfunctions closely related to ASD, thereby uncovering the underlying neurobiological mechanisms. VaeTF not only improves the classification performance of ASD but also provides a new method and theoretical support for objective assessment and early diagnosis based on fMRI.

Lien vers le texte intégral (Open Access ou abonnement)

4. John LP, Hemalatha R. Building Family Resilience in Autism Spectrum Disorder: An Integrative Review of Models, Interventions, and Future Directions. Ann Afr Med. 2026.

Families caring for children with autism spectrum disorder (ASD) often encounter persistent stress, stigma, and limited support systems. Despite these challenges, many demonstrate resilience-the capacity to adapt and sustain well-being under adversity. This integrative review synthesizes conceptual frameworks and interventions designed to strengthen family resilience in ASD caregiving. A systematic search across PubMed, Scopus, CINAHL, and Google Scholar (2000-2025) yielded 55 eligible studies. Seven resilience models were identified: protective factors, transactional, ecological systems, CARE, mindfulness-based stress reduction, Triple-A, and strengths-based approaches. Interventions informed by these models produced measurable benefits, including reductions in parental stress and caregiver burden, alongside gains in family cohesion, psychological health, and quality of life. The review underscores the importance of culturally adapted, multidisciplinary, and scalable resilience-building strategies, while calling for standardized evaluation tools and policy integration to ensure sustainable support for families raising children with ASD.

Lien vers le texte intégral (Open Access ou abonnement)

5. Lau DTY, Chan MMY, Mo FYM, Hung SF, Lai KYC, Leung PWL, Shea CKS. Cognitive Profiles in Adolescents and Young Adults With Co-Occurring Autism and First-Episode Psychosis: A Preliminary Neuropsychological Investigation. Psych J. 2026; 15(1): e70073.

Autism spectrum disorder (ASD) and psychosis are traditionally considered distinct psychiatric conditions with divergent developmental trajectories, yet emerging evidence suggests they may share overlapping neurodevelopmental characteristics. This study examined whether the cognitive profile associated with co-occurring autism and first-episode psychosis (FEP) reflects additive or interactive influences of the two conditions. Neuropsychological profiles were compared across four age-, sex-, intelligence quotient-, and education level-matched groups of adolescents and young adults (n = 45; aged 13-21): individuals with co-occurring ASD and FEP (FEP-ASD), FEP without ASD (FEP-O), ASD without FEP, and non-autistic controls. The FEP-ASD group exhibited an uneven cognitive profile characterised by relative strengths in visuospatial processing and recognition memory, alongside marked impairments in information processing speed, attentional control, and working memory. This pattern resembled the ASD profile but at a lower overall performance level, consistent with the additive impact of psychosis on ASD-related cognitive characteristics. FEP-ASD participants outperformed FEP-O in recognition memory, a domain usually preserved in ASD but impaired in psychosis. These preliminary findings suggest that co-occurring ASD and psychosis may produce a cognitive profile shaped by influences from both conditions. Larger longitudinal and multimodal studies are needed to clarify the underlying mechanisms.

Lien vers le texte intégral (Open Access ou abonnement)

6. Liu P, Liu S, Zhang J. Identifying the risk factors of autism spectrum disorders in infants born preterm: a systematic review and meta-analysis. Eur Child Adolesc Psychiatry. 2026.

Compared with infants born at term, infants born preterm are more likely to develop autism spectrum disorder (ASD), but specific risk factors remain uncertain. To identify possible risk factors associated with ASD development in infants born preterm. We systematically searched PubMed, Embase, Web of Science, CINAHL and Scopus to May 2025 for observational studies reporting ASD in infants born preterm. Meta-analyses used RevMan 5.4 and Stata 14.0. We assessed certainty of the evidence using the GRADE approach. A total of 45 studies were included in the systematic review, of which 40 were included in the meta-analyses. Twenty-two potential risk factors were identified. Prenatal factors included male, small for gestational age (SGA), lower birth weight, lower gestational age, low maternal education, ethnic minorities, antepartum hemorrhage (APH), and multiple births. Postnatal factors included severe cranial ultrasound abnormality, cerebral palsy (CP), severe retinopathy of prematurity (ROP), bronchopulmonary dysplasia (BPD), more days on mechanical ventilation, necrotizing enterocolitis (NEC), sepsis, longer neonatal intensive care unit (NICU) stay, delivery room resuscitation, longer duration of oxygen therapy, postnatal steroid administration, inotropic support, small head circumference (HC) at birth, and lack of breastfeeding. Overall, the certainty of evidence was low or very low, indicating that the associations between these risk factors and ASD in infants born preterm are highly uncertain. Our findings suggest that these 22 prenatal and postnatal factors may be associated with an increased risk of ASD in infants born preterm; however, due to low or very low certainty of evidence, the conclusions remain highly uncertain. Further high-quality prospective studies are needed to confirm these associations and to inform risk stratification and targeted preventive strategies.

Lien vers le texte intégral (Open Access ou abonnement)

7. Logrieco MG, Bertamini G, Casula L, Chetouani M, Guerrera S, Fasolo M, Venuti P, Scattoni ML, Fulceri F, Vicari S, Cohen D, Valeri G. Parental Stress and Caregiver Role Modulate Child-Caregiver Prosodic Synchrony in Autism: A Computational Analysis. Autism Res. 2026: e70189.

Parental stress influences parent-child interactions in typical development and is a prognostic factor of autism outcome. However, we still do not know to what extent parental stress affects parent-child interactions and whether caregiver role matters. This study explored the relationship between parental stress and prosodic synchrony in parent-child vocal interactions, drawing on complex dynamic systems and affective computing frameworks. We assessed 62 dyads (31 autistic preschoolers, interacting separately with their mother and father) during structured play interactions at two time points (12 months apart) along with perceived parental stress. We used a Deep Learning model to segment child-caregiver acoustic interactions with high accuracy automatically. Downstream, prosodic synchrony was modeled through cross-recurrence quantification analysis. Linear mixed-effects models were used to assess the impact of parental stress, caregiver role, and time on synchrony metrics. Models showed significant associations between parental stress and synchrony metrics for spectral and formant amplitude features. Higher stress levels were linked to less stable, predictable, and structured interactions. These effects were more pronounced in father-child dyads compared to mother-child dyads. Permutation analyses confirmed that these associations were specific to moment-to-moment coordination rather than general acoustic similarity. In autistic children, parental stress levels are linked with the temporal dynamics of parent-child prosodic synchrony, specifically affective speech and moment-to-moment coordination. It appears to be more pronounced in fathers. The results underscore the importance of fostering parental well-being and tailoring interventions to account for differences between maternal and paternal interaction patterns in autism. Parents of autistic children often experience high levels of stress, which can affect how they communicate with their child. This study found that when parents—especially fathers—feel more stressed, their communication with their child becomes less coordinated and more difficult to follow. Supporting parents’ mental health could help improve how they connect and communicate with their autistic children. The study employed artificial intelligence techniques to analyze parent–child vocal interactions. eng.

Lien vers le texte intégral (Open Access ou abonnement)

8. Mandal AS, Shinohara RT, Jung B, Gardner M, Akouri HE, Yerys BE, Low KJ, Cole TJ, Guthrie W, Janke KM, Herrington JD, Hocking MC, Ball G, Payne JM, North KN, Muhlert N, Garg S, Seidlitz J, Fisher MJ, Alexander-Bloch AF. NF1-Specific Growth Charts for Head Circumference Over the First 3 Years of Life. Neurology. 2026; 106(2): e214480.

BACKGROUND AND OBJECTIVES: Macrocephaly is among the most common findings in neurofibromatosis type 1 (NF1) and may be associated with other clinical manifestations of the genetic syndrome. NF1-specific growth charts that account for expected macrocephaly may increase sensitivity for detecting atypical growth. We aimed to produce NF1-specific growth charts of head circumference for the age range of 0-3 years and to assess their potential clinical impact. METHODS: Using electronic health records from the Children’s Hospital of Philadelphia, we collected head circumference measurements from children with NF1 and a community control cohort seen at scheduled well-child visits. We compared head circumference normalized using Centers for Disease Control (CDC) growth charts between these groups over time. We constructed NF1-specific growth charts using 2 independent methods. Finally, we used mixed-effects models to relate the resulting centile scores to developmental delay assessed with the Survey of Well-being of Young Children. RESULTS: Our data set contained 2,180 observations from 305 individuals (167 male) with NF1 and 104,750 observations from 16,742 individuals (8,809 male) in the community control cohort, all aged 0-3 years. Head circumference was significantly elevated in the NF1 cohort across the age range (p-adjusted <0.05), but the Cohen effect size d varied nonlinearly with age, starting moderate at 1 month (d = 0.56), then small at 4 months (d = 0.28), moderate again at 15 months (d = 0.58), and finally large at 28 months (d = 0.80). NF1-specific growth curves demonstrated slower increases in head circumference in the first 2 months of life, yet more sustained growth over time. Although none of the children with NF1 met the standard for microcephaly according to CDC charts, smaller head circumference benchmarked against NF1-specific charts was correlated with developmental delay (standardized β = 0.24; p < 0.02). DISCUSSION: We present the first NF1-specific growth charts for head circumference covering the age range of 0-3 years. Macrocephaly in NF1 becomes more exaggerated over time as rate of growth is sustained compared with controls. Smaller head size relative to NF1 growth expectations is not captured by CDC charts yet it nevertheless relates to developmental delay, suggesting that NF1-specific charts may increase sensitivity to clinically concerning patterns of growth in children with NF1.

Lien vers le texte intégral (Open Access ou abonnement)

9. Mortari Ferreira G, de Souza Silva CM, Sampaio Rodrigues Pereira A, Sousa Silva Bonasser L, do Nascimento Araújo MG, de Oliveira Barros M, Sousa Damasceno R, Negreiros F, Rodrigues da Silva IC. Cognitive Profile of Autism and Intellectual Disorder in Wechsler’s Scales: Meta-Analysis. Eur J Investig Health Psychol Educ. 2026; 16(1).

Autism spectrum disorder (ASD) and intellectual disability (ID) frequently coexist and share heterogeneous cognitive manifestations, yet their specific performance patterns on Wechsler scales remain poorly systematized. This meta-analysis synthesized data from 31 studies using the WISC-IV, WISC-V, WAIS-III, and WAIS-IV to compare cognitive index profiles in individuals with ASD, ID and ASD+ID. Standardized mean differences (Hedges’ g) were calculated using random-effects models, adopting a normative reference of mean 100 and SD 15. Results showed a distinct profile for ASD, with greater impairments in the Processing Speed Index (PSI) and Working Memory Index (WMI), while the Vocabulary Comprehension Index (VCI), Perceptual/Fluid Reasoning Index (PRI/FRI), and Visual Processing Index (VPI) remained close to normative scores. In contrast, ID and ASD+ID exhibited generalized deficits across all indices, with the lowest scores in Full-Scale IQ (FSIQ) and broad effects above g = -2.5. No significant differences emerged between Wechsler versions or age-based test types. Heterogeneity was high in ASD and ID across outcomes, but negligible in ASD+ID due to reduced k. These findings reinforce that ASD presents a specific cognitive pattern, whereas ID and ASD+ID display diffuse impairment, and that Wechsler scales are consistent across versions for identifying these profiles.

Lien vers le texte intégral (Open Access ou abonnement)

10. Osterne VJS, Oliveira MV, Pinto-Junior VR, Mota FSB, Cavada BS, Nascimento KS. Does Altered Membrane Glycosylation Contribute to Neurodevelopmental Dysfunction in Autism Spectrum Disorder?. Membranes (Basel). 2026; 16(1).

Neuronal development relies on cell-surface glycoconjugates that function as complex bioinformational codes. Recently, altered glycosylation has emerged as a central mechanistic theme in the pathophysiology of autism spectrum disorder (ASD). Critically, the brain maintains a distinctively restricted glycan profile through strict biosynthetic regulation, creating a specialized landscape highly susceptible to homeostatic perturbation. This « membrane-centric vulnerability » spans both glycoproteins and glycolipids; however, evidence remains fragmented, obscuring their pathogenic interplay. To bridge this gap, this review synthesizes evidence for these two primary classes of membrane glycoconjugates into a unified framework. We examine how defects in key glycoproteins (such as NCAM1 and neuroligins) directly impair synaptic signaling, trafficking, and plasticity. We then demonstrate how these defects are functionally coupled to the glycolipid (ganglioside) environment, which organizes the lipid raft platforms essential for glycoprotein function. We propose that these two systems are not independent but represent a final common pathway for diverse etiological drivers. Genetic variants (e.g., MAN2A2), environmental factors (e.g., valproic acid), and epigenetic dysregulation (e.g., miRNAs) all converge on this mechanism of impaired glycan maturation. This model elucidates how distinct upstream causes can produce a common downstream synaptic pathology by compromising the integrity of the membrane signaling platform.

Lien vers le texte intégral (Open Access ou abonnement)

11. Sertié AL, Josino R, Goll VR, Nunes Goussain Filippo AL, Campos GDS, do Rego F, Siqueira ES, Farias de Alcântara N, Zachi EC, Passos-Bueno MR. Catatonia and regression in an autism spectrum disorder patient harbouring a BRSK2 frameshift mutation. J Med Genet. 2026.

Deleterious variants in the BRSK2 gene, which encodes a serine/threonine kinase crucial for neuronal polarisation and brain development, have recently been linked to the pathogenesis of autism spectrum disorder (ASD). However, comprehensive clinical descriptions of individuals with pathogenic BRSK2 variants remain limited, and the molecular and cellular consequences of these mutations are poorly understood. This case report provides a detailed clinical, cognitive and molecular characterisation of a male patient with ASD harbouring a de novo BRSK2 frameshift variant, who developed catatonia, developmental regression and cognitive decline during early adolescence. To assess the functional impact of the variant, induced pluripotent stem cells (iPSCs) and iPSC-derived neural organoids were generated from the patient. Molecular analyses revealed a significant reduction in BRSK2 transcript and protein levels. Sequencing of BRSK2 mRNA showed exclusive expression from the wild-type allele, consistent with degradation of the mutant transcript via nonsense-mediated decay. These findings broaden the mutational and phenotypic spectrum associated with BRSK2-related neurodevelopmental disorders and provide functional evidence supporting the pathogenicity of the identified variant. Furthermore, this report demonstrates the role of BRSK2 in complex neuropsychiatric features-such as catatonia and cognitive deterioration, which remain underreported in the existing literature-and emphasises the importance of longitudinal cognitive and behavioural monitoring in individuals with BRSK2 mutations.

Lien vers le texte intégral (Open Access ou abonnement)

12. Sun Y, Chen H, Zhu Y, Li C, Jiang H, Jia Y. Intelligent monitoring-based screen exposure patterns and neurodevelopmental outcomes in preschool children. Eur J Pediatr. 2026; 185(2): 108.

Few studies have analyzed preschoolers’ screen exposure patterns, especially combined screen time and content, and the associations with neurodevelopment. This study aims to identify the screen exposure patterns in preschoolers by intelligent technology and to examine their associations with their neurodevelopment. This cross-sectional study enrolled preschool children from two kindergartens in Shanghai. Data were collected from March 2023 to July 2023. Screen time and content types were monitored over 7 consecutive days using an intelligent monitoring technology validated by the 24-h diary method (κ = 0.61). Neurodevelopmental outcomes were assessed using the Ages and Stages Questionnaire, Third Edition (ASQ-3); developmental abnormality was defined as a score < 1 SD from the mean in each domain. K-means clustering analysis identified screen exposure patterns, and binary logistic regression was applied to examine associations between screen exposure patterns and neurodevelopmental outcomes. Of 355 preschool children included, 204 were males (57.5%) and 251 (70.7%) were aged between 34.5 months and 50.5 months. K-means cluster analysis yielded 4 screen exposure patterns: restrictive use, moderately educational, noneducational, and educational-dominant pattern. Binary logistic regression showed the moderately educational pattern was associated with gross motor abnormalities (OR = 2.530, 95% CI: 1.089-5.875, P = 0.031), and non-educational pattern with fine motor abnormalities (OR = 3.172, 95% CI: 1.122-8.968, P = 0.029). CONCLUSION: This monitoring study identified heterogeneous screen exposure patterns in preschool-aged children, revealing that excessive use of moderately educational content and noneducational content was associated with lower gross motor and fine motor skills. When limiting total screen time, parents should focus on content selection for preschool-aged children. Future research should focus on the objective measurement of different types of screen content and utilize the intelligent monitoring system to conduct cohort studies, aiming to explore the causal associations between screen exposure content and children's development. WHAT IS KNOWN: • Few studies have analyzed preschoolers' screen exposure patterns (especially combined time and content) or the associations with neurodevelopment, with scarce research using objective measures of both. WHAT IS NEW: • Using validated intelligent monitoring, we identified heterogeneous patterns and found excessive moderately educational/noneducational content linked to lower motor skills.

Lien vers le texte intégral (Open Access ou abonnement)

13. Zheng B, Zhang N, Hu M, Zhou Y, Zhang Z, You Y, Xiao C, Zhao Q, Feng S, Wang X, Ping Y, Mo X, Chen J, Wang Y, Liu X, Zheng Y, Xu C, Lin HW, Wang H, Hu W, Chen Z. Prenatal glucocorticoids exposure disrupts maturation of a cluster of microglia, causing poor social memory and more repetitive behaviors. Cell Rep. 2026; 45(1): 116820.

Prenatal glucocorticoid exposure is associated with higher risks of autism spectrum disorder (ASD), yet the underlying mechanisms remain poorly understood. Here, we report that late-pregnancy exposure to dexamethasone (a synthetic glucocorticoid) induces social memory deficiency and increased repetitive behaviors in offspring with an imbalance of neurotransmission in the hippocampus. Single-cell RNA sequencing uncovers an expansion of MRC1(+) microglia, with arrested maturation. Strikingly, this population exhibits selective F13a1 (encoding a coagulation factor functioning as transglutaminase) upregulation. Early postnatal inhibition of F13A1 restores microglial maturation and ameliorates behavioral abnormalities. An elevated level of F13A1 is also observed in the plasma of postnatal rats and human umbilical cord blood exposed to prenatal glucocorticoids. Together, it suggests that prenatal glucocorticoid exposure disrupts the maturation of MRC1(+) microglia, thereby causing social memory impairment and increased repetitive behaviors. This underscores that arrested maturation within a cluster of microglia may be related to ASD and identifies F13A1 as a promising therapeutic target and biomarker.

Lien vers le texte intégral (Open Access ou abonnement)