Autism Spectrum Disorder

Literature and regulatory information updated: 4 October 2026.

Genetic architecture, syndromic forms, genetic assessment, family counselling and treatment

0. Initial Reference Table

1. Introduction

Autism spectrum disorder is a neurodevelopmental condition defined by persistent deficits in communication and social interaction across multiple contexts, associated with restricted, repetitive patterns of behaviour, interests or activities, with onset in the early developmental period and clinically significant functional impact [2, 5]. The concept stems from a historical line beginning with Leo Kanner, a psychiatrist at Johns Hopkins University, who in 1943 described eleven children with “autistic disturbances of affective contact”, characterised by extreme social isolation, obsessive insistence on sameness, echolalia and preservation of isolated cognitive abilities [29]. A year later, Hans Asperger in Vienna published a description of four boys with “autistic psychopathy”, with preserved formal language, circumscribed interests and difficulty with social reciprocity, a presentation that would become known as Asperger syndrome [30].

Genetic understanding of autism began with the study by Susan Folstein and Michael Rutter (1977), who, in 21 twin pairs, demonstrated concordance of 36% in monozygotic twins and 0% in dizygotic twins for strictly defined autism, with much greater concordance for a broader cognitive phenotype; this was the first robust argument against the psychogenic theory and in favour of a hereditary basis [31]. Subsequent research showed concordance of 60–90% in monozygotic twins and 20–30% in dizygotic twins, confirming heritability but also a contribution from shared environmental factors [4]. More recent heritability estimates, obtained in Nordic population cohorts and cohorts from five countries with more than two million individuals, are approximately 80% [7, 8].

The decisive conceptual transition occurred with DSM-5 (2013), which replaced the categories “autistic disorder”, “Asperger disorder”, “childhood disintegrative disorder” and “pervasive developmental disorder not otherwise specified” with a single designation — autism spectrum disorder — reduced the three deficit domains to two (communication and social interaction; restricted, repetitive patterns), included sensory abnormalities in the criteria for the first time, and introduced severity levels defined by support needs; it also created the separate category of social (pragmatic) communication disorder [4]. Anyone with a previous diagnosis within the spectrum, including Asperger syndrome, retains the diagnosis [4]. Research has not confirmed the validity of the former subcategories and instead points to multiple aetiologically distinct “autisms” [4].

From the perspective of medical genetics, ASD is now understood as a point of phenotypic convergence of hundreds of aetiologies. Only a minority of children with autism receive a specific genetic diagnosis, and almost all well-established aetiologies — fragile X syndrome, tuberous sclerosis complex, Rett syndrome and many others — are associated with intellectual disability [4]. Use of chromosomal microarray as a first-line investigation confirmed that rare CNVs (population frequency < 1%) are present in 5–10% of individuals with ASD, mainly involving genes of the neuronal synaptic complex [4]. From 2012, exome sequencing in simplex families demonstrated the role of de novo loss-of-function variants and the influence of paternal age, and large international consortia identified more than a hundred risk genes, organised into three main pathways: synaptic formation and function, transcriptional regulation and chromatin remodelling [4, 13, 14, 32]. In this article, ASD is treated as a clinical entity with heterogeneous genetic architecture, with dedicated subsections for the monogenic forms most relevant to the geneticist: SHANK3 (Phelan–McDermid syndrome), CHD8, PTEN (with macrocephaly) and SCN2A.

2. Molecular Genetics and Pathophysiology

2.1. Overview of genetic architecture

The genetic architecture of ASD combines four classes of variation with effects of very different magnitude. In the synthesis by Devlin and Scherer reproduced in the Oxford Desk Reference, ASD-associated syndromes (fragile X, tuberous sclerosis, untreated phenylketonuria, Rett, Timothy, Potocki–Lupski) contribute approximately 10% of cases; rare chromosomal abnormalities (maternal 15q11-q13 duplication, trisomy 21, Turner syndrome 45,X, 47,XYY and 47,XXY) approximately 5%; rare CNVs whose aetiological genes are not yet defined (16p11.2, dup 7q11.23, 22q11.2, 1q21.1, 15q13.3, dup 17p12) approximately 5%; and rare high-penetrance genes (PTCHD1/PTCHD1-AS, NRXN1, SHANK1, SHANK2, SHANK3, NLGN3, NLGN4X, NRXN3, CNTNAP2, DPP6) approximately 5%; the remainder corresponds to multigenic contributions of rare and common variants, intellectual disability and neuropsychiatric disease genes, environmental effects on gene expression and mutation rate (parental age), in proportions not yet determined [4]. The Brazilian reference text summarises this heterogeneity as multifactorial inheritance involving more than 400 genes and CNVs [1].

Source: adapted from Devlin and Scherer (2012) in the Oxford Desk Reference and subsequent genomic studies [4, 33, 34].

2.2. Common variation and the polygenic component

The study by Gaugler and colleagues in the Swedish population cohort estimated narrow-sense heritability at approximately 52%, mostly residing in common variation, whereas de novo mutations explained only 2.6% of liability variance [33]. The meta-analysis of genome-wide association studies (GWAS) by Grove and colleagues (18,381 individuals with ASD and 27,969 controls, based on the Danish iPSYCH cohort) identified the first five genome-wide significant loci and seven further loci shared with schizophrenia, depression and educational attainment, demonstrating quantitative and qualitative polygenic heterogeneity between ASD subtypes [34]. Common variation explains most familial aggregation, but each variant has a minimal effect (odds ratios of around 1.1), so polygenic scores have no individual diagnostic or predictive utility to date.

2.3. De novo variants and high-risk genes

Exome sequencing of 343 simplex families by Iossifov and colleagues (2012) showed an excess of de novo loss-of-function variants in genes linked to synaptic plasticity and revealed that small de novo insertions/deletions and substitutions are predominantly paternal in origin, with an increase dependent on the father's age [4]. This effect has a population basis: in Iceland, each additional year of paternal age adds approximately two de novo mutations to the offspring's genome [35]. Expanded analysis of the Simons Simplex Collection, involving approximately 2500 families, estimated that, including CNVs, coding de novo mutations contribute to approximately 30% of all simplex ASD diagnoses and 45% of diagnoses in females, and that genes affected in females overlap those affected in boys with a lower intelligence quotient (IQ) [32]. De Rubeis and colleagues (2014) showed that many implicated genes encode proteins involved in synaptic formation, transcriptional regulation and chromatin remodelling [4].

Subsequent studies integrating de novo variants, inherited variants and case–control data increased the number of risk genes. Satterstrom and colleagues (35,584 samples, 11,986 with ASD) identified 102 genes with a false discovery rate (FDR) ≤ 0.1, of which 49 had a higher frequency of de novo variants in individuals with severe neurodevelopmental delay and 53 in individuals with ASD; 13 are located in recurrent CNV regions, and expression is enriched in excitatory and inhibitory neuronal lineages, consistent with excitation–inhibition imbalance [14]. Fu and colleagues (63,237 individuals) identified 72 genes with FDR ≤ 0.001 (185 with FDR ≤ 0.05); de novo truncating variants, deleterious missense variants and CNVs accounted for 57.5%, 21.1% and 8.44% of association evidence, with CNVs conferring the highest relative risk; meta-analysis with developmental delay cohorts increased the number of genes associated with ASD or developmental delay to 373 [13]. Zhou and colleagues, in 42,607 cases (including 35,130 from the SPARK cohort), identified 60 exome-wide significant genes and five new moderate-risk genes (NAV3, ITSN1, MARK2, SCAF1, HNRNPUL2); carriers of loss-of-function variants in these moderate-risk genes had a lower frequency of cognitive impairment than carriers of variants in high-penetrance genes such as CHD8, SCN2A, ADNP, FOXP1 and SHANK3 (59% versus 88%) [36]. The SFARI Gene database, maintained by the Simons Foundation, aggregates and classifies evidence gene by gene and is the usual reference for consultation [37].

2.4. Copy number variants

CNVs identified in ASD frequently contain genes for synaptic cell adhesion molecules (neurexins, neuroligins, SHANK), supporting the hypothesis of a synaptogenesis defect [4]. The most frequent CNV is a deletion/duplication of approximately 600 kb at 16p11.2, present in approximately 0.5% of ASD cohorts; next are del Xp22.11, del 2p16.3 (NRXN1), del 7q11.23, del 22q11.2 (DiGeorge region), del 1q21.1 and del 15q13.3 [4]. The 16p11.2 deletions are more penetrant than the duplications, and penetrance estimates for clinical use have been published [4, 38]. Many of these CNVs are inherited from unaffected parents, indicating the coexistence of other genetic and environmental factors and preventing prenatal diagnosis of these CNVs from accurately predicting the development of ASD [4]. The SHANK family (three members) is implicated by rare CNVs in several phenotypes — schizophrenia, attention deficit hyperactivity disorder (ADHD) and intellectual disability — showing that similar biological pathways can give rise to different phenotypes [4]. In 22q11.2 deletion syndrome, the neuropsychiatric phenotype is age-dependent: ADHD (32%) and autism (21%) at younger ages, anxiety (35%) at all ages and schizophrenia spectrum disorders in 24% of young adults and in 41% of individuals over 25 years [39]. The reciprocal 22q11.2 microduplication, described in a total of more than 200 patients with a highly variable but generally milder phenotype, is associated with ASD in 14% to 25%; approximately 70% of duplications are inherited, usually from an unaffected parent [39]. Among sex chromosome aneuploidies, trisomy X (47,XXX) showed no increase in autism risk in the study by Bishop and colleagues (2011), although it is associated with a higher risk of ADHD/hyperactivity, which reached 50% in a parental survey (Tartaglia and colleagues, 2012) [40].

2.5. High-risk monogenic forms: gene-by-gene distinctions

SHANK3 (22q13.33) — SHANK3-related Phelan–McDermid syndrome. SHANK3 encodes a scaffold protein of the excitatory postsynaptic density that links NMDA receptors and metabotropic glutamate receptors, through GKAP/PSD-95 and Homer complexes, to the actin cytoskeleton, controlling formation, maturation and maintenance of dendritic spines and AMPA- and mGluR-mediated transmission [16]. Haploinsufficiency — caused by 22q13.3 deletion (from < 50 kb to > 9 Mb), an intragenic variant or a rearrangement disrupting the gene — destabilises the postsynaptic density; the syndrome is considered a paradigm of synaptopathy [16]. Between 69% and 74% of terminal deletions occur on the paternal chromosome 22 [16]. Durand and colleagues described the first SHANK3 variants in individuals with ASD in 2007 [41]. Larger deletions correlate with more severe phenotypes and smaller ones with ASD; small deletions involving SHANK3 may have reduced penetrance and variable expressivity, and SHANK3 variants have been described in apparently non-syndromic autism and schizophrenia [16, 42].

CHD8 (14q11.2) — CHD8-related neurodevelopmental disorder with overgrowth. CHD8 encodes a chromodomain DNA-binding helicase, an ATP-dependent chromatin remodeller that acts as a transcriptional repressor (histone H1 recruitment, suppression of p53-mediated apoptosis) and negative regulator of the Wnt/β-catenin pathway [17]. The mechanism is heterozygous loss of function, usually de novo; the gene is highly intolerant of loss-of-function variants in the general population [17]. Bernier and colleagues (2014), resequencing CHD8 in 3730 children with developmental delay or ASD, identified 15 independent mutations, no truncating variant in 8792 controls, and defined an autism subtype with macrocephaly, characteristic facial features and gastrointestinal complaints [43]. In mouse and zebrafish models, macrocephaly results from hyperproliferation of progenitors in the frontal cortex [17].

PTEN (10q23.31) — PTEN hamartoma tumour syndrome with ASD and macrocephaly. PTEN encodes a lipid phosphatase that removes the D3 phosphate from phosphatidylinositol-3,4,5-trisphosphate and antagonises the PI3K-AKT-mTOR pathway (mammalian target of rapamycin); germline haploinsufficiency leads to unopposed AKT/mTOR signalling, providing the basis for studying mTOR inhibitors [18]. In large series, missense variants are overrepresented in individuals with ASD, and individuals with PTEN-related ASD have functionally less severe variants than individuals with PHTS without ASD [18].

SCN2A (2q24.3) — SCN2A-related disorders. SCN2A encodes the α subunit of the voltage-gated sodium channel Nav1.2. Variants associated with epilepsy beginning before 3 months predominantly produce gain of function (slowed fast inactivation, accelerated recovery or increased persistent current), whereas variants associated with later-onset epilepsy and the group with ASD and intellectual disability predominantly produce loss of function; truncating variants occurred exclusively in the late-onset group [44]. This functional dichotomy has direct therapeutic consequences (section 7).

2.6. Integrated pathophysiology

Convergence of risk genes on three processes — synapses, transcriptional regulation and chromatin — and their early expression in the fetal cortex support the current model: disturbances during neurogenesis and cortical circuit maturation alter excitation–inhibition balance and connectivity, with effects on social and sensory processing [5, 13, 14]. Genes with stronger evidence in ASD than developmental delay are enriched in maturing neurons and overlap schizophrenia genes, whereas genes associated with developmental delay are enriched in progenitors and immature neurons [13]. Excess brain growth with early macrocephaly was considered typical of ASD for years, but this association was challenged when appropriate population norms were used (Raznahan et al., 2013) [4]; macrocephaly nevertheless retains aetiological relevance in monogenic forms (PTEN, CHD8), in which it is constant or almost constant [17, 18]. In the series by Bolton and colleagues (2001), idiopathic infantile macrocephaly was associated with increased ASD risk (odds ratio 5.44) [45].

2.7. Environmental factors

Prenatal exposure to sodium valproate is associated with an absolute ASD risk of 4.42% (adjusted hazard ratio 2.9) and a childhood autism risk of 2.50% (adjusted hazard ratio 5.2) in the Danish cohort of 655,615 children [46]. Advanced parental age, extreme prematurity and congenital infections are other associated factors; population studies do not support a contribution from relevant non-genetic maternal effects (0.4–1.6% of variance) [5, 7]. In thalidomide embryopathy, the neurological manifestations include facial palsies and an increased incidence of autism and epilepsy [47]. In the spectrum of defects induced by maternal hyperthermia in pregnancy, in addition to neural tube defects and craniofacial anomalies, ID, ASD, ADHD and hypotonia have been described; the risk of autism associated with fever was lower in mothers who took antipyretics [48]. In September 2025, the FDA announced the initiation of a paracetamol label change to reflect a possible association between use during pregnancy and neurodevelopmental disorders, acknowledging in the same announcement that causality had not been established and that contradictory studies exist [49]; the following day, the EMA stated that there was no new evidence warranting changes to European recommendations, maintaining paracetamol as an option for pain or fever during pregnancy when clinically necessary [50]. In the Swedish cohort of approximately 2.48 million children, the marginal association with ASD in the population analysis (hazard ratio 1.05; 95% CI 1.02–1.08; absolute risk difference at 10 years 0.09%) disappeared in the sibling comparison (hazard ratio 0.98; 95% CI 0.93–1.04), pointing to familial confounding rather than a causal effect [51]. A regulatory announcement is not equivalent to a demonstration of causality, and the message for pregnant women should follow the recommendations in force in the relevant jurisdiction. The 2026 Haute Autorité de Santé (HAS) guideline reaffirms that there is no evidence of a link between ASD and vaccination (including the measles, mumps and rubella vaccine), paracetamol use in pregnancy or overexposure to screens, although the latter does not remove the need for rules of balanced use [28].

3. Clinical Spectrum and Manifestations by System

3.1. Core behavioural features

The two core domains are persistent deficits in communication and social interaction and restricted, repetitive patterns of behaviour, interests or activities [1, 2]. The first domain includes deficits in socioemotional reciprocity (from atypical social approaches and difficulty maintaining conversation to reduced sharing of interests and emotions), deficits in non-verbal communicative behaviours (eye contact, body language, gestures and facial expression) and deficits in developing, maintaining and understanding relationships (adjusting behaviour to context, sharing imaginative play, making friends) [1]. The second domain includes stereotyped or repetitive motor movements, use of objects or speech, insistence on sameness with inflexible adherence to routines, fixed and restricted interests abnormal in intensity or focus, and hyper- or hyporeactivity to sensory stimuli [1]. Language delay is no longer a mandatory criterion: it is not universal, particularly in children with intelligence in the normal range, but language pragmatics (saying things inappropriate to or disconnected from the conversation, use of unusual vocabulary) are always impaired [4].

The cognitive profile is heterogeneous: there is no typical pattern, verbal abilities may either exceed or fall below non-verbal abilities, and some children have unusual isolated abilities relative to their overall level [4]. In US surveillance in 2022, 39.6% of 8-year-old children with ASD and available cognitive assessment had intellectual disability (ID), varying between population groups (52.8% in Black children and 32.7% in non-Hispanic White children) [10]. The global median of ASD with co-occurring ID is 33% [9].

3.2. Expression across the lifespan and severity levels

Source: Síndromes Genéticas: Reconhecer para Tratar, Table 7.1.1 [1].

Source: adapted from Table 7.1.2 of Síndromes Genéticas: Reconhecer para Tratar and DSM-5-TR [1, 2].

The 2026 HAS guideline adopts the notion of profound autism, proposed by the 2022 Lancet Commission, for children and adolescents with intensive and continuous support needs, 24 hours a day, every day of the year [28]. It is defined by three features: severe disorder of intellectual development, with marked learning difficulties; pronounced limitation of adaptive functioning, requiring constant supervision for safety and for activities of daily living; and absent or majorly limited receptive and expressive communication [28]. The intensity of needs becomes evident over time, not at the initial diagnosis, especially when the response to early interventions is insufficient or there is developmental regression; it should not be confused with severe challenging behaviours, nor call into question the diagnosis, early intervention, schooling or access to services [28]. The recommendations for this group include reinforced time structuring, augmentative and alternative communication (AAC) and sensory adaptations, maintenance of acquired skills and of autonomy in daily living, regular surveillance of somatic (particularly epilepsy) and psychiatric comorbidities — more frequent and harder to identify, because pain is rarely expressed explicitly —, use of the solutions developed for profound multiple disabilities and reinforced support for parents and siblings, including respite solutions [28].

3.3. Associated manifestations by system

Nervous system. Epilepsy risk increases in adolescence, but epidemiological data indicate that this risk is concentrated in individuals with significant ID (Viscidi et al., 2014) [4]; in the minority subgroup with severe ID, up to 20% develop seizures by adulthood, often beginning in adolescence [4]. In children with developmental disorder and macrocephaly, seizure risk is higher (odds ratio 7.7 in the Israeli community study by Nevo and colleagues) [45]. Motor abnormalities (hypotonia, dyspraxia, atypical gait) and sleep disturbances are frequent [5].

Mental health. ADHD, anxiety disorders, obsessive–compulsive disorder, depression and, in a minority, psychotic disorders co-occur and may require psychopharmacological treatment [4, 52]. From adolescence onwards, the frequency of depression and anxiety warrants active screening for warning signs and referral to child and adolescent psychiatry or adult psychiatry; suicidal ideation and plans should be explored directly, particularly in the presence of psychiatric comorbidity, gender incongruence, or a history of bullying or physical or sexual abuse [28]. The same guideline stresses that there is no robust evidence of an association between autism and gender incongruence or dysphoria and recommends, when they coexist, a cautious approach and collaboration between the autism and gender identity teams [28].

Digestive system and nutrition. Constipation, abdominal pain, extreme food selectivity and diarrhoea are common and may manifest solely as behavioural change [53]. Bruxism is frequent and warrants dental follow-up [1]. Food selectivity can cause nutritional deficiencies and requires global assessment (including oral health) and behavioural and dietary management; bowel function should be monitored regularly because of the risk of constipation [28].

Pain. Pain may be expressed atypically, through self-injury or aggression towards others or through increased withdrawal, and should be sought with observational scales adapted to the mode of communication, such as the pain assessment grid for intellectual disability (GED-DI) or the simplified pain assessment scale for people with communication difficulties (ESDDA); gynaecological pain, including menstrual pain, deserves specific attention [28]. Any unusual change in behaviour calls for prompt physical examination in search of a medical cause [28].

Growth. Head circumference (HC), height and weight should always be measured; extreme macrocephaly (> 2.5–3 standard deviations [SD]) points to PTEN and generalised overgrowth to CHD8 or other overgrowth syndromes [4, 15].

3.4. Specific phenotypes of monogenic forms

SHANK3-related Phelan–McDermid syndrome. This is characterised by neonatal hypotonia, absent or severely delayed speech (receptive language is consistently better than expressive language), global delay and moderate to profound ID, with mild dysmorphic features (long eyelashes, pointed chin, broad or bulbous nose, periorbital fullness, large fleshy hands, dysplastic toenails); normal stature and HC distinguish it from other autosomal chromosomal abnormalities [16]. Between half and three quarters of individuals meet ASD criteria (approximately 60% in the GeneReviews frequency table), with chewing of non-food objects, reduced pain perception, teeth grinding and a tendency towards hyperthermia due to hypohidrosis; lifetime epilepsy prevalence exceeds 60% [16]. Regression differs from that observed in idiopathic ASD and Rett syndrome in being later and predominantly affecting motor and independence skills: in a study of 42 individuals (4–48 years), parents reported regression in 43%, beginning around 6 years, language loss in 33% (mean age 3 years), motor skill loss in 50% and independence skill loss in 50% (mean age 4 years), and recovery in approximately 40% over a period of 1 month to 10 years [16]. Neuropsychiatric decompensation (catatonia, bipolar disorder) emerges in adolescence or adulthood and may worsen at puberty, particularly in females [16]. Renal abnormalities (10–40% in deletions > 1 Mb) and lymphoedema accompany larger deletions rather than intragenic variants [16]. Ring chromosome 22 carries a risk of NF2-related schwannomatosis (16% of 44 individuals), caused by mitotic ring loss followed by somatic inactivation of the remaining NF2 allele, and atypical teratoid/rhabdoid tumours have been described [16, 54].

CHD8-related neurodevelopmental disorder with overgrowth. Based on 115 reported individuals, frequencies are: macrocephaly 80% (sometimes at birth, but often developing or becoming more pronounced in the first year), tall stature 80% (particularly at puberty), ASD 75–80% (moderate mean severity on ADOS-2, Autism Diagnostic Observation Schedule, 2nd edition), developmental delay/ID 75–80% (ID usually mild to moderate; regression in up to half), sleep disturbance 67% (difficulty falling asleep and nocturnal awakenings), gastrointestinal problems 63% (mainly constipation, with or without diarrhoea), ADHD 50%, anxiety 29%, hypotonia 27% and seizures 12% [17]. Approximately 80% of individuals with ASD also have ID [17]. Facial features include a prominent supraorbital ridge, broad forehead, hypertelorism, downslanting palpebral fissures, pointed chin and large or posteriorly rotated ears; childhood-onset dystonia (responding to deep brain stimulation in two cases), Chiari I malformation, psychosis, schizophrenia and catatonia in adults have been described [17]. The boy:girl ratio is approximately 2:1 (67% male) [17]. There is no evidence of increased cancer risk [17].

PTEN — ASD with macrocephaly. In childhood, PHTS manifests with macrocephaly caused by megalencephaly, almost universal (94%) and often exceeding +5 standard deviations, more pronounced than in idiopathic ASD [18]. Approximately half of children with PHTS have a neurodevelopmental disorder (ASD and/or developmental delay); in a large study, ASD, developmental delay or ID were present in 35% of children and 23% of adults [18, 55]. The neurocognitive profile includes language delay, deficits in working memory and processing speed, and adaptive and sensory abnormalities; ASD severity is usually in the mild to moderate range [18]. Psychiatric disorders (anxiety, depression, bipolar disorder, ADHD, obsessive–compulsive disorder, psychosis) affect 11–44% and epilepsy 6–17%, with greater risk in individuals with ASD; epilepsy may be associated with focal cortical dysplasia, polymicrogyria or cavernomas [18]. In children, pigmented macules on the glans, lipomas, vascular malformations, linear naevi, thyroid nodules (Hashimoto thyroiditis and thyroid carcinoma from 4 years) and gastrointestinal polyps (25–30% in a paediatric series) should be sought [18]. In the phenotype historically designated Bannayan–Riley–Ruvalcaba syndrome, birth weight exceeds 4 kg, mean HC is +4.89 standard deviations, and there are hypotonia and motor and language delay (50%), ID (15–20%), seizures (25%), ileocolic hamartomatous polyps (45%), lipomas (75%), proximal myopathy (60%) and pigmented penile spots that may emerge only in late childhood [56]. In adulthood, lifetime cancer risk, according to the longitudinal analysis by Yehia and colleagues, is 91% for female breast cancer, approximately 48% for endometrial cancer, 33% for epithelial thyroid cancer, 30% for renal cell carcinoma, 17% for colorectal cancer and 5% for cutaneous melanoma [18, 57]. The Oxford Desk Reference, with earlier data, reports 85% for breast cancer (50% by 50 years), 35% for thyroid cancer (median 37 years, follicular histology overrepresented, never medullary), 28% for endometrial cancer, 35% for renal cancer (predominantly papillary), 9% for colon cancer and > 5% for melanoma, as well as Lhermitte–Duclos disease in up to 32% [58].

SCN2A-related disorders. The spectrum includes, at the benign end, self-limited familial neonatal–infantile seizures; at the severe end, developmental and epileptic encephalopathy 11 (Ohtahara syndrome, epilepsy of infancy with migrating focal seizures, West syndrome); forms beginning after 3 months (myoclonic–atonic epilepsy, Lennox–Gastaut syndrome, focal epilepsies with continuous spike-and-wave during sleep); and a growing group with ASD and ID, with rare or absent seizures; episodic ataxia type 9 and chorea are additional associations [44].

4. Clinical Recognition and Warning Signs

Early recognition rests on developmental surveillance and universal screening. The American Academy of Pediatrics (AAP) recommends specific ASD screening at 18 and 24 months, in addition to surveillance at every visit, and immediate referral for diagnostic assessment and intervention when ASD is suspected, without waiting for confirmation [53]. Warning signs include no response to name at 12 months, absence of pointing for sharing or joint attention, reduced eye contact, absence of symbolic play at 18 months, loss of words or social skills at any age, intense restricted interests and unusual sensory reactions [5, 53]. The 2026 French guideline converges: intervention starts at the first warning signs, without waiting for diagnostic confirmation, and excessive screen use reported in the history should not be considered the cause of the social communication difficulties nor delay assessment [28]. Diagnosis should be accessible at any age; in the adolescent or young adult with late suspicion, a multidimensional and multidisciplinary assessment is recommended, to avoid both overdiagnosis and underdiagnosis [28].

In the genetics clinic, clinical history should include a three-generation pedigree exploring parental and close relatives' educational attainment, occupational choices, hobbies and sociability; prenatal, perinatal and postnatal history; detailed developmental history, particularly language; and specific enquiry about regression or loss of skills [4]. Complete physical examination should focus on minor anomalies, neurological findings, social interaction and behaviour, with skin examination under a Wood lamp for hypopigmented macules (tuberous sclerosis) when there is a seizure history, and recording of HC, height and weight [4].

Signs increasing the probability of a monogenic or chromosomal aetiology and guiding investigation are as follows:

In tuberous sclerosis, children with seizures — particularly infantile spasms — in the first two years of life have a greater probability of learning difficulties and ASD than those developing seizures later or remaining seizure-free [4]. In assessing an adult with thyroid carcinoma, macrocephaly or benign tumours should raise the possibility of PHTS [58].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Clinical diagnosis

Diagnosis is clinical and requires fulfilment of criteria A (three communication and social interaction deficits), B (at least two of four restricted, repetitive patterns), C (symptoms present in the early developmental period, potentially becoming evident only when social demands exceed capacities or being masked by learned strategies), D (clinically significant impairment) and E (not better explained by ID or global developmental delay; when they co-occur, social communication must be below that expected for developmental level) [1, 2]. Assessment by a child psychiatrist or clinical psychologist includes observation of social interaction (eye contact, greeting, turn-taking), repetitive activities, play (imaginative or otherwise), peer relationships and responses to relatives, strangers and new situations [4]. The most widely used standardised instruments are ADOS-2 (Autism Diagnostic Observation Schedule, 2nd edition) and ADI-R (Autism Diagnostic Interview-Revised); M-CHAT-R/F (Modified Checklist for Autism in Toddlers, Revised with Follow-Up) is the most widely used screening instrument [5, 53]. The median age of the earliest documented diagnosis in the USA in 2022 was 47 months [10].

The 2026 HAS guideline separates the diagnostic assessment, the first stage of intervention, from the functional assessment, which accompanies the whole trajectory: the latter is multidimensional and multidisciplinary (medical, psychological, allied health and educational), carried out at least once a year by trained professionals and followed by a synthesis meeting, and always includes a physical examination, structured clinical observations and continuous informal observation by parents and professionals [28]. The domains to be reassessed, organised according to the International Classification of Functioning, Disability and Health (ICF), are communication and language, social interaction, cognitive functioning, sensory and motor functioning, emotions and behaviour, somatic health, autonomy in daily living, school and pre-vocational learning, and the family and material environment [28]. Speech and language therapy assessment, including pragmatics, forms part of the diagnostic and functional assessment, and written language should be assessed from 6 years of age even when speech is limited; interventions may begin before all initial assessments are completed [28].

Source: tables 3 to 9 of the 2026 HAS guideline [28].

5.2. Genetic assessment: ACMG recommendations

The 2013 American College of Medical Genetics and Genomics (ACMG) guideline defines the geneticist's roles as establishing the aetiology where possible, counselling and contributing to management, and recommends: confirming the ASD diagnosis before any investigation; discussing options, yield and costs; coordinating with the treating physician; recognising the expanded phenotypes of known syndromes; and defining an individualised plan [15]. It proposes a tiered approach: the first tier comprises a three-generation pedigree, dysmorphology assessment, CMA for all, FMR1 testing in all boys (and in girls with a suggestive phenotype) and MECP2 testing in girls; the second comprises PTEN sequencing when HC exceeds +2.5 standard deviations, brain magnetic resonance imaging for specific signs (regression, seizures, microcephaly, neurological abnormalities) and metabolic investigation guided by clinical signs [15]. The AAP adopted an equivalent approach in 2020, recommending CMA and fragile X testing in all children with ASD and accepting ES as an investigation to consider when the initial tests are negative [53]. The 2026 French guideline likewise recommends systematically offering a genetics clinic consultation as part of the aetiological assessment of children and adolescents with ASD [28].

The 2021 ACMG guideline strongly recommends considering exome or genome sequencing (ES/GS) as a first- or second-line investigation in paediatric patients with congenital abnormalities beginning before 1 year or developmental delay or ID beginning before 18 years, based on clinical utility, a higher diagnostic rate than conventional investigations and potential cost-effectiveness when requested early [59]. This guideline does not explicitly include isolated ASD without ID; however, the meta-analysis and consensus by Srivastava and colleagues (2019) showed an ES yield of 36% in neurodevelopmental disorders, 31% in isolated forms and 53% in forms with associated conditions, exceeding the 15–20% yield of CMA, and proposed ES as a first-line investigation [20]. The 2010 consensus had already established CMA as a first-line investigation for developmental delay, ID, ASD and multiple congenital abnormalities, with a yield of 15–20% [60].

The Oxford Desk Reference summarises British practice: microarray in children with ID and/or dysmorphic features; whole-exome or whole-genome sequencing (WES/WGS) with low yield in non-syndromic ASD and relevant yield in patients with additional signs suggesting monogenic disease; fragile X testing if developmental delay or ID is present; consideration of MECP2 in girls with normal early development followed by regression; non-routine neuroimaging, reserved for specific neurological signs; targeted electroencephalography (EEG) or EEG in the presence of the triad of severe learning difficulty, autism and epilepsy (Baird et al., 2003) [4].

5.3. Investigation yield

5.4. Practical algorithm

The proposed algorithm, integrating ACMG, AAP and Oxford Desk Reference recommendations, follows these steps: (1) confirmation of the clinical ASD diagnosis by a specialist team; (2) clinical genetic assessment with a three-generation family history, dysmorphology examination and measurement of HC, height and weight, including parental HC adjusted for height; (3) clinical suspicion of a specific syndrome — targeted testing (for example, TSC1/TSC2, MECP2, PTEN); (4) no specific suspicion — CMA and FMR1 testing in parallel or ES/GS as the initial investigation where available, particularly if ID, congenital abnormalities or dysmorphic features are present; (5) non-diagnostic CMA — trio ES (with CNV analysis where possible) or an ASD/ID/overgrowth panel; (6) HC > +2.5 SD — ensure PTEN analysis by sequencing and deletion/duplication testing; (7) terminal 22q13.3 deletion on CMA — mandatory karyotype to exclude ring chromosome 22; (8) result with a variant of uncertain significance — familial segregation, periodic reassessment and no use in clinical or reproductive decisions [4, 15, 16, 18, 53].

5.5. Particular aspects of molecular diagnosis by gene

In Phelan–McDermid syndrome, most cases result from 22q13.3 deletions undetectable by SHANK3 sequencing, so CMA is usually the first investigation; intragenic variants, undetectable by CMA, require sequencing; ID panels are appropriate only if they include deletion/duplication testing; if CMA shows a terminal deletion, karyotyping is mandatory to investigate ring chromosome 22, because this finding changes surveillance (NF2-related schwannomatosis) [16]. In CHD8-related disorder, single-gene sequencing is rarely useful and is not recommended; CMA, multigene ID/ASD/overgrowth panels and ES/GS are preferred [17]. In PHTS, diagnosis requires a heterozygous pathogenic or likely pathogenic PTEN variant; the optimal strategy includes sequencing of exons 1–9 and flanking intronic regions, followed by deletion/duplication analysis and, in negative cases with a Cowden phenotype, investigation of the PTEN promoter and other susceptibility genes (SDHx, PIK3CA, AKT1, KLLN epimutations) [18, 58]. Approximately 10% of cases with a Bannayan–Riley–Ruvalcaba phenotype without a coding sequence variant have large deletions [58]. Cowden clinical criteria (pathognomonic, major and minor) and the Cleveland Clinic risk calculator aid patient selection; in children, macrocephaly associated with at least one of ASD or developmental delay, dermatological lesions (lipomas, trichilemmomas, oral papillomas, penile freckling), vascular lesions (arteriovenous malformations, haemangiomas) or gastrointestinal polyps should prompt consideration of PHTS [58]. In SCN2A-related disorders, patch-clamp electrophysiology defines gain or loss of function, but is performed in a minority; a prediction tool based on alignment of functionally characterised variants in the nine sodium channel genes exists [63].

5.6. Limitations of genetic tests and report content

CMA does not detect point variants, balanced rearrangements (including the parental translocation giving rise to 22q13.3 deletion), ring chromosomes or low-level mosaicism, and its resolution varies by platform [4, 16, 60]. ES does not reliably detect FMR1 CGG expansion, methylation and imprinting abnormalities (for example, maternal 15q11-q13 duplication requires confirmation of parental origin), deep intronic and regulatory variants or low-level mosaicism, and has variable sensitivity for small CNVs; GS improves detection of CNVs and non-coding variants, but fragile X still requires repeat PCR with methylation analysis [4, 15, 59]. In PHTS, sequencing alone misses large deletions and promoter variants, so test coverage should include PTEN deletion/duplication analysis [18, 58]. A negative result does not exclude a genetic cause in a convincing phenotype: what the test covered should be checked before abandoning the hypothesis.

The report should state the gene, reference transcript with version, Human Genome Variation Society (HGVS) nomenclature at coding DNA and protein levels (or coordinates and gene content for CNVs), zygosity, ACMG/AMP classification with the criteria applied and technical limitations [64]; CNVs are classified according to the joint ACMG/ClinGen standard [65]. Variants should be normalised to the MANE transcript and as in ClinVar [66]: for the genes in this article, MANE Select (v1.5) is NM_001372044.2 for SHANK3, NM_001170629.2 for CHD8, NM_000314.8 for PTEN, NM_001040142.2 for SCN2A (with NM_001371246.1 as MANE Plus Clinical), NM_002024.6 for FMR1, NM_000548.5 for TSC2 and NM_001110792.2 for MECP2 (whose literature predominantly uses the historical numbering of isoform e2, NM_004992.4, now MANE Plus Clinical). The SHANK3 MANE transcript is recent and earlier literature uses other transcripts, so correspondence should be confirmed before comparing variants. When the literature uses older notation, both forms are stated: the PTEN variants reported by Butler and colleagues as H93R, D252G and F241S are written NP_000305.3:p.(His93Arg), p.(Asp252Gly) and p.(Phe241Ser) [61]. A variant of uncertain significance does not become pathogenic because of a treatment response or partial segregation in a family (many ASD variants are inherited from parents with a mild or absent phenotype); segregation in these circumstances is research to clarify the variant rather than predictive testing, and a variant of uncertain significance should not form the basis of prenatal or preimplantation diagnosis [64].

6. Differential Diagnosis

Differential diagnosis has two aspects: distinguishing ASD from other disorders with overlapping symptoms, and identifying the underlying aetiology.

In the context of macrocephaly, causes of an apparently enlarged head (hairstyle, craniosynostosis) should be excluded, comparison made with parental HC adjusted for height, and it should be remembered that PTEN and PTCH1 variants may present with few syndromic signs, so benign familial macrocephaly should be diagnosed cautiously [45]. In Proteus syndrome, which may mimic PHTS, the cause is somatic AKT1 mosaicism; older cases of “Proteus with PTEN” are now considered PHTS [58].

7. Treatments and Risk Management Strategies

7.1. General principles

Educational and behavioural interventions are the basis of treatment [4]. No drug is approved for the core social communication symptoms; drugs target specific behaviours (irritability, aggression, self-injury, hyperactivity, insomnia) and comorbidities [52, 53]. The current emphasis has shifted from assigning a label to identifying all the individual's needs and their impact on life, with the extent of support required guiding the plan [4]. NICE (National Institute for Health and Care Excellence) guidance for those under 19 recommends a psychosocial intervention focused on social communication, mediated by parents or carers, and advises against antipsychotics, antidepressants, anticonvulsants, exclusion diets, secretin, chelation and hyperbaric oxygen for the core symptoms [72]. The 2026 French guideline, which updates that of 2012, maintains the principles of early, global and coordinated intervention, based on educational, behavioural and developmental approaches, centred on a personalised plan co-constructed with the family, with functional objectives by domain, resources, responsible persons and review deadlines; if there is no progress after several months, the approach, intensity and objectives should be reviewed with the family and, if necessary, with a specialist team or an autism resource centre [28].

7.2. Early intervention and non-pharmacological therapies

The randomised Early Start Denver Model trial (48 children aged 18 to 30 months, 2 years of intervention) showed improvement in IQ, adaptive behaviour and diagnosis compared with community intervention [73]. The PACT (Preschool Autism Communication Trial) trial of parent-mediated social communication showed a sustained reduction in overall autism symptom severity at long-term follow-up (approximately 6 years after intervention ended) [74]. Applied behaviour analysis (ABA), individualised and usually delivered individually with a certified analyst, is recommended in GeneReviews for children with Phelan–McDermid syndrome or PHTS who meet ASD criteria [16, 18]. The multidisciplinary plan described in the Brazilian reference text includes essential stimulation (improvement of overall neurological development), speech and language therapy (fonoaudiologia in Brazil: eye contact, relationships, auditory comprehension, non-verbal imitation, vocal play, verbal imitation, expressive and communicative speech), psychotherapy (emotional control, behaviour modification and problem-solving), occupational therapy (feeding, dressing, hygiene, continence), comprehensive paediatric care, dentistry (bruxism), music therapy, play therapy and equine-assisted therapy (complementary therapy involving horses), the latter with limited evidence [1]. In non-verbal individuals, augmentative and alternative communication is essential, particularly in Phelan–McDermid syndrome given the discrepancy between receptive and expressive language [16].

The HAS recommends starting early intervention as soon as the first warning signs appear, without waiting for diagnostic confirmation, and offering infants and children with a confirmed risk or a diagnosis of ASD at least 10 hours per week of intervention by professionals trained in developmental and behavioural approaches and in parent coaching, in addition to parent-mediated activities at home and the work of early childhood and school professionals [28]. Before 6 years of age, it gives priority to interventions with evidence of efficacy on the developmental trajectory — early intensive behavioural intervention (EIBI, based on ABA principles), naturalistic developmental behavioural interventions (NDBI, such as ESDM and JASPER), developmental interventions (PACT) and parent-mediated interventions — and to inclusion in mainstream early childhood settings and preschool education [28]. The table summarises the meta-analysis on which the guideline is based; cells with a dash correspond to absent data or non-significant results [28].

HAS grades: B, scientific presumption (low-powered randomised trials, cohort studies); C, low level of evidence. Source: table 2 of the 2026 HAS guideline [28].

From 6 to 19 years, the HAS recommends social skills training groups, effective in children and adolescents without ID (grade B); from 6 to 12 years, cognitive behavioural therapy for social communication (grade C) and parent-mediated behavioural interventions to reduce challenging behaviours (grade B); progress is measured with standardised instruments, at least annually and more often in early childhood and at transitions [28]. The reference programmes differ in age and prerequisites: ABA and TEACCH apply at all ages and without prerequisites; JASPER from 12 months to 8 years, in children with limited language; ESDM starts between 1 and 3 years and continues up to 5, requires a developmental age of at least 7–9 months and ceases to be appropriate if this exceeds 4 years in all domains; PACT, from 12 months up to a maximum of 10 years, consists of a 1 to 1.5 h session every two weeks, with video analysis, and 30 minutes of daily parent play at home [28].

All plans include a personalised communication component, receptive and expressive, without prerequisites; AAC, with visual, gestural and digital supports (for example, tablets), should be introduced early when indicated (grade C), supervised by a trained professional and generalised to all settings, with parent training; learning to write should not be ruled out in a non-speaking child aged 6 years or older [28]. Social skills are worked on from 6 years of age, in groups or individually and in natural settings, with video modelling when appropriate [28]. Sensory assessment (hyper- and hyposensitivity, seeking or avoidance) and motor assessment guide environmental adaptations; between 0 and 2 years, in the presence of severe hypotonia, poor postural adjustment or weak social engagement, standardised assessment and early intervention in psychomotor therapy, physiotherapy or occupational therapy are recommended, and from 3 years a physical or sporting activity integrated into the programme and task-oriented motor interventions [28]. The environment should be made predictable and accessible — spaces with differentiated functions and quiet areas, structured but flexible schedules, clear and visual information —, and teleconsultation only replaces face-to-face care when there is no alternative, with documented consent and regular reassessment [28].

Support for parents is a mandatory component of the care pathway: parent coaching from the time of suspicion (for example, PACT and JASPER), psychoeducation, parenting skills programmes, therapeutic education, carer training, peer support groups and respite solutions, with attention to siblings' distress and prevention of parental burnout; parental involvement in global interventions is recommended with grade B [28]. Parents should be warned about interventions presented as curative or that demand exclusivity, because interrupting treatment of comorbidities represents a lost opportunity [28]. Regular physical activity is part of the plan; animal-assisted, music or art mediation may be offered according to the child's interests but, without validated evidence, does not constitute therapy nor replace established interventions [28].

7.3. Pharmacological treatment of irritability and aggression

For challenging behaviours, the HAS proposes a sequential algorithm: the core symptoms (social communication difficulties, restricted and repetitive interests and activities) have no pharmacological indication; in the presence of aggression towards others or self-injury, irritability or emotional dysregulation, environmental causes (relational, sensory, contextual) are sought first, with functional analysis and adjustment of the environment; without relief, somatic and pain assessment follows (ear, nose and throat, gastrointestinal, oral and dental, urinary); without a treatable cause, a specialist medical consultation in search of chronic disease (inflammatory, autoimmune, metabolic, neurological or the behavioural phenotype of a syndrome); and, without a diagnosis, a specialist or emergency psychiatric consultation [28]. For prevention, a consistent AAC system across settings when speech is scarce (grade C), adaptation of the environment and anticipation of triggers are recommended; for treatment, behavioural psychoeducational techniques (grade C) with positive reinforcement and, in combination, certain drugs; for self-injury, protective equipment such as a helmet may be considered [28]. Educational or pharmacological intervention may proceed in parallel with medical investigation, but does not replace it [28].

Risperidone has shown benefit in treating tantrums, aggression and self-injury in children with these features [4]. In the pivotal RUPP (Research Units on Pediatric Psychopharmacology) network trial, 101 children (5–17 years, mean age 8.8 years) were randomised to risperidone (0.5–3.5 mg/day) or placebo for 8 weeks; the Aberrant Behavior Checklist irritability score (ABC-I) decreased by 56.9% versus 14.1%, response rates were 69% versus 12%, mean weight gain was 2.7 kg versus 0.8 kg, and benefit was maintained at 6 months in two thirds of responders [75]. Aripiprazole was assessed in two 8-week trials (2–15 mg/day) in patients aged 6 to 17 years, with a significant ABC-I reduction compared with placebo [23, 76]. The Brazilian PCDT of 2022, based on a systematic review, estimated a mean difference for risperidone of −8.31 ABC-I points (95% CI −11.45 to −5.17), additional short-term weight gain of 1.75 kg and a serum prolactin increase of 27.42 ng/mL compared with placebo; it found no difference between aripiprazole and risperidone (low certainty) and does not recommend cannabidiol [25].

Management of adverse effects includes baseline and periodic measurement of weight, body mass index, waist circumference, blood glucose and lipid profile, surveillance for extrapyramidal symptoms and tardive dyskinesia, and enquiry about hyperprolactinaemia symptoms (gynaecomastia, galactorrhoea, menstrual changes) [22, 53]. The PCDT recommends identifying the target behaviour, starting with a low dose and using the minimum effective dose, reviewing efficacy and adverse effects at 3–4 weeks and stopping treatment if there is no clinically significant response after 6 weeks; after a sustained response, gradual dose reduction is considered [25]. In adults, the PCDT specifies risperidone starting at 1 mg/day (preferably at night), increases of 1 mg every 3–4 days and a usual therapeutic dose of approximately 3 mg/day (1–6 mg/day), with a maximum of 10 mg/day [25]. In the European Union, aripiprazole is authorised only for schizophrenia (from 15 years) and manic episodes in bipolar I disorder (from 13 years), without an ASD indication, and the harmonised paediatric indication for risperidone (European Commission decision of 7 October 2008) concerns short-term symptomatic treatment (up to 6 weeks) of persistent aggression in conduct disorder in children aged 5 years and over and adolescents with below-average intellectual functioning or ID, rather than ASD [79]; use in this population is therefore off-label, and NICE guidance allows antipsychotics for challenging behaviour only when psychosocial interventions are insufficient [72, 78, 80].

The 2026 French guideline does not set antipsychotic doses but defines prescribing rules: psychotropic drugs should be exceptional and temporary; initial prescription falls to a child and adolescent psychiatrist, psychiatrist, paediatrician, paediatric neurologist or neurologist, except in emergencies, with information to the general practitioner; they require a prior physical assessment with appropriate complementary tests, monotherapy, starting at a low dose with progressive titration, assessment at each renewal and reassessment of the indication at least every 6 months (or earlier, according to the marketing authorisation), including a treatment break or deprescribing; ineffectiveness or poor tolerance lead to discontinuation; and multiple prescribers and anticipatory prescriptions should be avoided [28]. Parents may refuse medication, unless there is a serious risk to the child's health, and admission to an institution cannot be made conditional on taking psychotropic drugs; in adolescence, the need for contraception should be considered [28]. Medical and dental procedures are prepared with progressive familiarisation, visual supports and practice visits; procedural pain is prevented with distraction, topical anaesthetic and an equimolar mixture of oxygen and nitrous oxide, and temporary paediatric restraint is only acceptable when safety is compromised and alternatives have failed, with medical prescription and presence and informed consent [28].

7.4. Insomnia, ADHD and other comorbidities

Paediatric prolonged-release melatonin (Slenyto®) was authorised in the EU on 20 September 2018 for insomnia in children and adolescents aged 2 to 18 years with ASD and/or Smith–Magenis syndrome, when sleep hygiene measures are insufficient. The Commission decision of 26 August 2024 extended the indication to neurogenetic disorders with aberrant daytime melatonin secretion and/or nocturnal awakenings, and that of 17 March 2025 added insomnia from 6 to 17 years with ADHD, under the same conditions [24, 26]. It is supplied as 1 mg and 5 mg tablets; the initial dose is 2 mg, increasing to 5 mg if the response is inadequate, to a maximum of 10 mg, once daily, 30 to 60 minutes before bedtime, with or after food; the effect should be assessed after at least 3 months and reassessed at least every 6 months; the most frequent adverse effects (1–10%) are somnolence, fatigue, mood changes, headache, irritability, aggression and a hangover sensation [24]. In the pivotal trial (125 children aged 2 to 17.5 years), prolonged-release melatonin increased total sleep time and reduced sleep latency compared with placebo [81]. Co-occurring ADHD is treated with psychostimulants, atomoxetine or α2-adrenergic agonists, with generally lower responses and more adverse effects than in isolated ADHD [52]. Depression and psychotic disorders may emerge and require specific psychopharmacological treatment [4].

The HAS places melatonin within a sleep algorithm: systematic screening (questionnaires and sleep diaries), search for medical causes (pain) and comorbidities (ADHD, restless legs syndrome, parasomnias, sleep apnoea, epilepsy), and sleep hygiene with cognitive behavioural therapy for insomnia as first line [28]. If these measures are ineffective or insufficient and the disorder has a relevant impact, prolonged-release melatonin 2 mg is added, 30 minutes before bedtime, for 15 days, with the goal of a sleep latency below 30 minutes and at most one nocturnal awakening; if this is not achieved, the dose is raised to 5 mg for 15 days and then to 10 mg for 15 days, and lack of response at 10 mg leads to a specialist sleep consultation [28]. The guideline accepts immediate- or prolonged-release formulations adapted to each child, calls for standardised initiation schedules and maximum doses, and recommends continuing melatonin after 18 years if the chronic disorder persists and has responded [28].

7.5. Management guided by genetic aetiology

Phelan–McDermid syndrome. Management is multidisciplinary, according to the updated 2023 consensus guidelines [54]. After diagnosis, renal ultrasonography (with voiding cystourethrography if indicated), echocardiography, ophthalmological and audiological assessment, gastrointestinal and feeding assessment, thyroid function and neurological assessment are recommended; surveillance includes annual audiology and ophthalmology, dental assessment and puberty monitoring [16]. Two points are mechanism-specific: given reduced pain perception, any unexplained behavioural change should prompt investigation for a painful cause (fractures, otitis, reflux); and individuals with ring chromosome 22 should have annual neurological examination by a physician experienced in NF2, annual brain magnetic resonance imaging from 10–12 years until the fourth decade, and annual audiology with auditory brainstem evoked potentials, avoiding radiotherapy for NF2-associated tumours [16]. High temperatures and prolonged sun exposure (hypohidrosis) and sharp objects or tight clothing (hypoalgesia) should be avoided [16]. In regression with suspected autoimmune encephalitis, a response to intravenous immunoglobulin has been described in four girls [16]. Among investigational treatments, insulin-like growth factor 1 (IGF-1) showed improved social interaction, restricted behaviour, sensory reactivity and hyperactivity in very small placebo-controlled crossover trials (9 and 10 children); lithium reversed catatonia and regression in two adolescents and stabilised atypical bipolar disorder in an adult [16].

CHD8-related disorder. Management is symptomatic: behavioural and/or pharmacological intervention for sleep disturbance, standard treatment of developmental delay, seizures and constipation, and surgical treatment of Chiari I malformation when indicated [17]. Brain magnetic resonance imaging at diagnosis is recommended if HC is ≥ +3 SD, is increasing rapidly or there are signs of cerebrospinal fluid obstruction or compression of the brainstem, cerebellum or cranial nerves, and may be considered above +2 SD without other symptoms; surveillance at each consultation includes growth parameters, development, anxiety, ADHD, aggression or self-injury, sleep and constipation [17]. There are no cancer surveillance recommendations [17].

PHTS with ASD. Identification of a PTEN variant in a child with ASD transforms follow-up into a lifelong cancer surveillance programme. According to the July 2026 GeneReviews update: annual physical and dermatological examination; annual education about cancer signs from 18 years (or 5 years before the earliest cancer in the family); annual thyroid examination and ultrasonography from 12 years; monthly breast self-examination from 18 years, clinical breast examination every 6–12 months from 25 years and annual mammography (with consideration of tomosynthesis) and contrast-enhanced breast magnetic resonance imaging from 30 years; renal imaging every 1–2 years from 40 years; assessment of endometrial cancer symptoms with consideration of biopsy every 1–2 years from 30–35 years; colonoscopy every 5 years from 35 years; annual monitoring of development and educational needs in childhood and adolescence, with ASD screening from 12 months; annual assessment of scoliosis and asymmetry [18]. For a suspicious thyroid, thyroidectomy should be total; skin lesions should be excised only if malignancy is suspected or relevant symptoms occur, because of the risk of keloids and rapid recurrence; unnecessary ionising radiation should be avoided [18]. Recommendations of the international consensus group for neurological and neurodevelopmental manifestations of PHTS were published in 2025 [55]. The ClinGen (Clinical Genome Resource) actionability group classified PHTS as strongly actionable in the adult context and moderately actionable in the paediatric context [82]. The placebo-controlled crossover trial of everolimus for neurocognitive outcomes (NCT02991807) showed good tolerability but no significant improvement in most measures; mTOR inhibitors should be restricted to clinical trials [18].

SCN2A-related disorders. In early-onset forms (before 3 months), usually with gain of function, sodium channel blockers (phenytoin, carbamazepine, oxcarbazepine) were associated with clinically relevant seizure reduction or seizure freedom; in later-onset forms, usually with loss of function, these drugs were rarely effective and sometimes worsened seizures, so should not be used empirically [44]. This rule is opposite to that for Dravet syndrome (SCN1A). Antisense oligonucleotides and CRISPR activation are experimental [44].

7.6. Interventions without evidence

Gluten- and casein-exclusion diets, chelation, hyperbaric oxygen and secretin have no demonstrated efficacy for the core symptoms, and some carry risks [53, 72]. Controlled trials of drugs targeting the core symptoms have also failed: intranasal oxytocin, balovaptan (a vasopressin V1a receptor antagonist) and bumetanide were not superior to placebo (section 7.7) [83–86]. The 2026 HAS guideline reserves exclusion diets (gluten-free or casein-free) for confirmed food intolerance or gastrointestinal disease, such as coeliac disease, under medical supervision, and attributes no specific indication in autism to food supplements [28]. It further advises against, owing to the absence of studies, the Doman–Delacato, Feuerstein and Padovan (neurofunctional reorganisation) methods, facilitated communication and related methods (such as the Rapid Prompting Method) and Saccade®; owing to insufficient evidence, primitive reflex integration, Son-Rise®, the 3i method, neurofeedback, psychoanalysis and psychoanalytically based interventions and Snoezelen; and, owing to studies demonstrating absence of effect, “auditory integration” practices, including the Tomatis method (grade B), and packing (wrapping in cold wet sheets) [28].

Leucovorin (folinic acid) requires careful distinction. On 22 September 2025, the FDA announced initiation of approval of calcium leucovorin for cerebral folate deficiency, presented as a cause of developmental delay with autistic features [87]; the label change actually approved in March 2026 restricts the indication to cerebral folate transport deficiency with a confirmed FOLR1 variant, based on 46 patients described in published reports and without a clinical trial, and does not include ASD or cerebral folate deficiency attributed to folate receptor autoantibodies [27]. In ASD without a FOLR1 variant, leucovorin is off-label use in the USA and all jurisdictions; randomised evidence is scarce, and one of the most cited controlled trials (Panda and colleagues, 2024) was retracted in January 2026 [88]. The American Academy of Pediatrics published guidance for paediatricians faced with increased demand, and leucovorin use after the FDA announcement was analysed in a publication in 2026 [89, 90]. For the geneticist, the practical implication is to investigate FOLR1 variants (and measure cerebrospinal fluid 5-methyltetrahydrofolate when the presentation is suggestive: regression, epilepsy, movement abnormalities, hypomyelination) rather than empirically prescribing leucovorin.

7.7. Pivotal trials and negative trials

Positive trials concern target behaviours (irritability, insomnia), rather than core symptoms, for which all large controlled trials were negative [75, 76, 81, 83, 84, 86]. Indirect comparison of risperidone and aripiprazole from these trials is not legitimate: populations, doses, magnitude of placebo response and ABC-I reporting (percentage reduction in RUPP, mean change in points in aripiprazole trials) differ; the Brazilian PCDT systematic review found no difference between them (low certainty) [25]. Relative reductions should be read alongside residual severity — in RUPP, approximately one third did not respond and metabolic effects were consistent — and the mean improvement of approximately half an hour of sleep with melatonin is a group mean, with variable individual response [75, 81]. The placebo response was large in trials of core symptoms, reinforcing the need for controlled trials before adopting therapies promoted outside them [83, 84].

7.8. Comparative regulatory status: United States and European Union

The hierarchy of evidence should be explicit: a peer-reviewed article, a sponsor announcement, an opinion of the EMA Committee for Medicinal Products for Human Use (CHMP), an authorisation decision (European Commission or FDA) and a public announcement of regulatory intent are different levels; a favourable opinion is not authorisation, and authorisation does not guarantee marketing, reimbursement or access in each country — in Portugal these depend on marketing and INFARMED funding assessment, and in Brazil on ANVISA registration and incorporation into SUS after CONITEC assessment (as occurred with risperidone in the 2022 PCDT) [25]. No medicinal product is authorised for core ASD symptoms in any jurisdiction. The table summarises the situation on 4 October 2026.

Distinguishing extrapolation from off-label use is essential in ASD: using risperidone in a 4-year-old child, or aripiprazole in the EU, is off-label use defined by label wording, age and jurisdiction, even if trial evidence in similar populations exists; prescribing prolonged-release melatonin for a child with ASD in the EU is within the indication, but the same prescription using a supplement in the USA lacks the quality assurance of an authorised medicinal product [22–24, 79].

7.9. Divergences between the 2026 HAS guideline and other sources

8. Reproductive Options and Family Counselling

8.1. Empirical recurrence risk

In the absence of an aetiological diagnosis, counselling is based on empirical risks. The prospective 2011 Baby Siblings Research Consortium study (664 infants) found that 18.7% of younger siblings of children with ASD developed the disorder, exceeding earlier estimates of 3–14%; male sibling sex (25.9% versus 9.6%; 2.8-fold increase) and more than one affected sibling (32.2% in multiplex families versus 13.5% in simplex families) were independent predictors [4, 92]. The 2024 update from the same consortium, with 1605 infants followed to 3 years at 18 centres, estimated recurrence of 20.2%, without a significant difference from 2011, and confirmed male sibling sex, more than one affected sibling and female proband sex (higher recurrence among siblings of affected girls) as predictors [21]. In population studies, risk is approximately 10 times that of the population in children with an autistic sibling [21, 93]. The Brazilian reference text proposes a risk of 3–10% for a couple with one affected child, higher if the affected child is female (approximately 7%) than male (approximately 4%), and 33–50% if two or more children are affected [1]. The difference between infant sibling study values and classical values reflects more sensitive ascertainment methods and inclusion of milder phenotypes [4]. These data were collected before systematic integration of CMA and ES, and a combined analysis would better inform counselling [4].

8.2. Recurrence when the cause is known

When a genetic cause is identified, risk becomes specific. For CNVs, parents should be tested; a parental balanced rearrangement (for example, translocation) increases risk [1]. For de novo variants in high-risk genes (CHD8, SCN2A, SHANK3, ADNP), sibling risk is low but above population risk because of possible gonadal mosaicism; carriers have a 50% transmission risk in each pregnancy [16, 17]. In Phelan–McDermid syndrome, approximately half of deletions resulting from unbalanced structural rearrangements have a parent carrying a balanced translocation; a parent with a non-mosaic deletion entails a 50% risk; parental mosaicism entails an increased but unquantifiable risk; and transmission of an intragenic SHANK3 deletion from a mother without ID or ASD to five affected daughters has been described, illustrating reduced penetrance [16]. In PHTS, 10–44% of cases are de novo and 10–50% have an affected parent, so careful parental examination (HC, skin, thyroid, breast) is indispensable; cascade testing of at-risk relatives is recommended, often in early childhood, and relatives without the variant have population risk [18, 58]. In fragile X syndrome, testing the mother and maternal lineage identifies premutation carriers with expansion risk and their own risk of primary ovarian insufficiency and tremor/ataxia syndrome [15].

8.3. Prenatal and preimplantation diagnosis

Prenatal diagnosis is not possible without a specific genetic diagnosis (for example, tuberous sclerosis or fragile X); prenatal use of CNVs of uncertain significance or determination of male fetal sex are insufficiently discriminatory for clinical use [4]. With a known familial variant, prenatal diagnosis and preimplantation genetic testing are possible, but the result does not predict phenotype in conditions with variable expressivity, such as Phelan–McDermid syndrome or incompletely penetrant CNVs [16, 38]. Advanced paternal age increases the number of de novo mutations and should be discussed as a population risk factor, without individual predictive utility [4, 35].

8.4. Communication with the family

Counselling should include variability of expression, even among close relatives, and published CNV penetrance estimates [4]. Discussion of broad phenotype traits in parents and siblings, explored in the pedigree, requires sensitivity [4]. Patient associations — the National Autistic Society in the United Kingdom, Federação Portuguesa de Autismo in Portugal, Autism-Europe at European level and the Phelan-McDermid Syndrome Foundation — are resources for families [4, 16, 94]. The relevant European Reference Network for syndromic forms and those with ID is ERN ITHACA (rare malformation syndromes, intellectual disability and other neurodevelopmental disorders); forms with tumour risk (PHTS) also fall within ERN GENTURIS [95, 96].

8.5. Organisation of care in France

In France, the care pathway is organised at three levels: identification in primary care, with a dedicated consultation; coordination and guidance platforms for neurodevelopmental disorders (PCO-TND, created in 2019 following the 3rd Autism Plan, 2013–2017); and diagnostic and tertiary teams — autism resource centres (CRA), grouped within the GNCRA, reference centres for language and learning disorders and specialist consultations in paediatric neurology, child and adolescent psychiatry and physical and rehabilitation medicine [28]. The National Strategy for Autism within Neurodevelopmental Disorders (2018–2022) created five centres of excellence — STRAS&ND (Strasbourg), iMIND (Lyon), CeAND (Montpellier), EXAC-T (Tours) and InovAND (Paris) — and a scientific interest group, and the National Strategy for Neurodevelopmental Disorders 2023–2027, coordinated by the Interministerial Delegation (DI-TND), continued it [28]. Intervention relies on early medico-social action centres (CAMSP), medico-psychological centres (CMP), special education and home care services (SESSAD) and medico-educational institutes (IME); the departmental house for persons with disabilities (MDPH) and its commission (CDAPH) decide the personalised compensation plan and the personalised schooling project (PPS), binding on the National Education system, under the Law of 11 February 2005 [28]. School arrangements include autism-specific preschool and primary teaching units (UEMA, with admission in the year of the 3rd birthday, and UEEA), localised units for school inclusion (ULIS) and support workers for pupils with disabilities (AESH), with continuity and a limited number of support workers [28]. The guideline recommends a care pathway coordinator designated on a long-term basis, territorial service directories published by the health insurance funds, the CRA, the PCO and the MDPH, tele-expertise from autism specialists for primary care, healthcare services adapted to disability (Handiconsult, Handisoins, Handident) and tools such as SantéBD; in child protection, it recommends training of professionals and foster families and access by services and judges to the official list of experts in ASD, ADHD and learning disorders [28]. The family associations consulted in drafting the guideline include Autisme France, Sésame Autisme and UNAPEI [28].

9. Special Considerations: Population Diversity and Epidemiology

Measured prevalence increased from approximately 6/1000 (Fombonne, 2003) to 1/88 (Centers for Disease Control and Prevention, CDC, 2012) and to 1/31 in 2022 in the USA, an increase attributed mainly to the broadening of criteria, greater awareness and identification capacity [4, 10]. ADDM (Autism and Developmental Disabilities Monitoring) surveillance in 2022, across 16 sites, estimated 32.2 per 1000 children aged 8 years, ranging from 9.7 (Laredo, Texas) to 53.1 (California); prevalence was lower in non-Hispanic White children (27.7) than in Asian or Pacific Islander (38.2), American Indian or Alaska Native (37.5), Black (36.6) and Hispanic children (33.0); the boy:girl ratio was 3.4; and cumulative incidence by 48 months was 1.7 times higher in children born in 2018 than in those born in 2014 [10]. The 2022 global systematic review (99 estimates from 71 studies) found a median prevalence of 100/10,000 (range 1.09–436/10,000), with a median boy:girl ratio of 4.2 and co-occurring ID in 33% [9]. The traditional sex ratio of 4:1 has been revised, with estimates close to 2.5:1 in population studies (Kim et al., 2011), suggesting underdiagnosis in females [4]. The 2026 French guideline places prevalence between 1% and 2% of the population, with considerable variation between countries and studies [28].

In Europe, the Portuguese national study of 2007 (332,808 school-age children on the mainland and 10,910 in the Azores) estimated 9.2/10,000 on the mainland and 15.6/10,000 in the Azores, with regional differences and associated medical conditions in 20%, including an unexpected frequency of mitochondrial respiratory chain disorders [11]. In Brazil, the Atibaia (São Paulo) pilot study estimated 27.2/10,000 (95% CI 17.6–36.8), with late diagnosis and limited service access [12]; the Brazilian reference text cites 70/10,000 [1]. No recent population prevalence studies with comparable methodology and national representativeness for Portugal, Brazil, most African and Middle Eastern countries or much of Asia were identified in the sources consulted, so these estimates should not be extrapolated; absence from the sources consulted does not prove absence of studies [9].

Genetically, most large exome and genome-wide association studies were performed in populations of European ancestry, reducing variant interpretation capacity in other ancestries and increasing the frequency of variants of uncertain significance in underrepresented populations [13, 34]. No founder effect has been described for ASD as an entity; in monogenic forms, almost all variants are de novo and private. In US surveillance, Black, Hispanic and Asian children with ASD had a higher frequency of co-occurring ID, probably reflecting inequalities in identification of forms without ID [10]. In the 2024 sibling cohort, race and maternal educational attainment were associated with recurrence risk, which the authors relate to social inequalities [21]. PHTS frequency, estimated in genotyped research populations, is approximately 1/9000 in All of Us (USA) and 1/13,000 in UK Biobank, higher than historical clinical estimates of 1/200,000 [18, 58, 97].

10. Illustrative Clinical Case

10.1. Published case: ASD with macrocephaly and a PTEN variant

Moreno Fernandes and colleagues, from a Portuguese hospital centre, described a 4-year-old girl with congenital hypothyroidism, progressive macrocephaly and global developmental delay, subsequently diagnosed with ASD [98]. Brain magnetic resonance imaging showed megalencephaly with prominent extra-axial spaces and a diffusely thickened corpus callosum, and genetic testing identified a novel frameshift variant in PTEN [98]. The authors emphasise that PHTS should be suspected in a child with autism and macrocephaly before oncological manifestations appear, enabling early diagnosis and appropriate surveillance [98]. The heterozygous variant was described as NM_000314.8:c.302del, p.(Ile101ThrfsTer12) (authors' notation: p.(Ile101Thrfs*12)), classified as likely pathogenic because it is a predicted loss-of-function variant in a gene with established haploinsufficiency; fragile X testing and CGH-array were normal, and the parents had not been tested at publication, so de novo origin has not been demonstrated [98]. The open-access article includes the HC curve (figure 1) and sagittal MRI with corpus callosum thickening (figure 2), available in the original version (https://pmc.ncbi.nlm.nih.gov/articles/PMC12871087/) [98]. Absence of parental testing does not prevent classification, because this rests on variant type; however, it affects counselling, which should include parental examination and testing, and the cancer surveillance summarised by the authors follows published guidelines rather than their own recommendation [18, 98]. This case reproduces the original observation by Butler and colleagues (2005), who found germline missense PTEN variants (H93R, D252G, F241S) in 3 of 18 children with ASD and macrocephaly, all with the largest HCs in the series [61].

10.2. Teaching case: approach in the genetics clinic

The following case is a composite, constructed for educational purposes from features described in GeneReviews, and does not correspond to a real patient.

A 3-year-old boy is referred for ASD (level 2) diagnosed at 30 months, with expressive language delay and no regression. His mother is 34 years old and father 41; a 6-year-old sister has typical development. On examination, HC is +3.4 SD, height +2.3 SD and weight at the 75th centile; he has a broad forehead, prominent supraorbital ridge and downslanting palpebral fissures; there are no hypomelanotic macules or pigmented glans macules. The family reports chronic constipation and frequent nocturnal awakenings. Maternal HC adjusted for height is normal; the father's HC is at the 90th centile, without skin lesions, and there is no family history of breast, thyroid or endometrial cancer. CMA and FMR1 testing are normal. Given macrocephaly with overgrowth, trio ES with CNV analysis is requested, identifying a de novo loss-of-function CHD8 variant classified as pathogenic according to ACMG and Association for Molecular Pathology criteria (ACMG/AMP). Counselling explains the haploinsufficiency mechanism, the 75–80% frequency of ASD and ID, absence of increased cancer risk, the need for brain magnetic resonance imaging (HC ≥ +3 SD) to exclude Chiari I malformation, and the plan for surveillance of sleep, constipation, anxiety and ADHD; recurrence risk for future siblings is low but not zero because of possible gonadal mosaicism, and prenatal or preimplantation diagnosis is possible; the risk for the proband's offspring will be 50% [17]. Prolonged-release melatonin is started at 2 mg at night, with reassessment at 3 months [24]. If ES had revealed a PTEN variant, the plan would include annual thyroid ultrasonography from 12 years and cascade testing of the parents, despite their being asymptomatic [18].

11. Prognosis and Long-Term Follow-Up

Autism was traditionally considered a lifelong, severely disabling disorder, but recognition in people with intelligence in the normal range has shown that prognosis is not always so limited [4]. Considerable uncertainty exists about long-term prognosis in ASD without ID, accounting for most current diagnoses in the USA and Western Europe; the most affected individuals may be unable to live independently, and prognosis is expected to improve with support in schools, higher education and employment [4]. The best prognostic factors are absence of ID, absence of language impairment and absence of mental comorbidity [1]. In the minority subgroup with severe ID, up to 20% develop epilepsy by adulthood, often beginning in adolescence [4]. In adolescence, most improve behaviour, and less impaired adults may function independently, although they remain socially vulnerable and prone to anxiety and depression [1]. Mortality exceeds that of the general population, with contributions from epilepsy, accidents and suicide [5].

Follow-up should be coordinated by a multidisciplinary team and adapted to aetiology: in idiopathic ASD, surveillance of development, sleep, behaviour, feeding, mental health and antipsychotic metabolic effects; during transition to adulthood, educational, occupational and social support planning [52, 53]. In monogenic forms, specific plans apply: lifelong cancer surveillance in PHTS, NF2 surveillance in ring chromosome 22, neuropsychiatric surveillance in adolescence in Phelan–McDermid syndrome (risk of catatonia and bipolar disorder), and epilepsy follow-up in SCN2A-related disorders [16, 18, 44]. Periodic reanalysis of exome data in undiagnosed cases, in the light of newly described genes, increases cumulative diagnostic yield [20].

The 2026 HAS guideline adds a follow-up plan in primary and specialist care common to all children and adolescents with ASD, which is added to the specific plans for monogenic forms [28]. For transition, it recommends anticipating adult life from 14–15 years, directing learning and administrative procedures towards adult priorities from 16 years, liaising early with adult services between 16 and 18 years, transmitting a written report of the care pathway with the young person's agreement, and maintaining AAC [28]. Relationship, emotional and sexual education should begin before puberty and continue through adolescence, with adapted content (puberty, menstruation, consent) and identification of signs of vulnerability, maltreatment or abuse; victims of sexual violence should receive adapted multidisciplinary support [28]. Transitions (entry into adolescence, change of school or team, family changes) require heightened surveillance and preparation with prior visits and progressive integration, and intervention teams should receive supervision preferably monthly and at least quarterly [28].

12. Summary and Final Considerations

Autism spectrum disorder is a clinically defined neurodevelopmental condition, with a prevalence between 1% and 3% at school age in populations with active surveillance, heritability of approximately 80%, and a genetic architecture that includes common polygenic variation (the largest fraction of heritability), rare CNVs (5–10%), rare de novo variants in more than a hundred genes and recognisable monogenic syndromes. High-risk genes converge on synaptic function, transcriptional regulation and chromatin remodelling. For the geneticist, four models illustrate gene-by-gene distinctions: SHANK3, with moderate to profound ID, hypoalgesia, late regression and the need for karyotyping to exclude ring chromosome 22; CHD8, with macrocephaly, tall stature, sleep disturbance and constipation; PTEN, with extreme macrocephaly and a lifetime cancer risk requiring surveillance from childhood; and SCN2A, in which the functional mechanism guides antiepileptic selection. The recommended genetic assessment combines dysmorphology assessment with CMA and FMR1 testing (ACMG 2013), MECP2 testing in girls with regression, PTEN testing when HC exceeds +2.5 SD, and ES/GS as a first- or second-tier investigation when ID, congenital anomalies or developmental delay are present (ACMG 2021), with a combined yield of approximately 16% in unselected ASD and higher in complex forms. The empirical sibling recurrence risk is approximately 20% in prospective studies, higher in male siblings, in multiplex families and when the proband is female, and becomes specific when the aetiology is known. Treatment rests on early educational and behavioural interventions; risperidone (0.5–3 mg/day) and aripiprazole (5–15 mg/day) are FDA-approved for associated irritability, with mandatory metabolic monitoring, and prolonged-release melatonin is EMA-approved for insomnia. Identification of a genetic cause changes surveillance, reproductive counselling and, in some cases, treatment, which justifies systematically offering genetic assessment to all people with ASD. In the last 12 months, the most relevant regulatory development — the US approval of leucovorin — was restricted to cerebral folate transport deficiency caused by a FOLR1 variant, which reinforces the role of genetic diagnosis; controlled trials of therapies targeting the core symptoms remain negative [27, 86]. The 2026 French guideline adds an intensity threshold for early intervention (at least 10 h/week), a melatonin titration scheme in 15-day steps, restrictive rules for psychotropic drugs and the notion of profound autism, and reaffirms the absence of a link with vaccines, paracetamol in pregnancy and screens [28].

13. References

1. Loiola JCR, Alexandre AF, Prata VPS, Ribeiro EM. Autismo ou transtorno do espectro autista (TEA). In: Ribeiro EM, coordinator. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 142-5.
2. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed., text revision (DSM-5-TR). Washington, DC: American Psychiatric Association Publishing; 2022. doi:10.1176/appi.books.9780890425787.
3. Orphanet. Autism (ORPHA:106) and Rare disease with autism (ORPHA:180772). Paris: INSERM; [accessed 3 October 2026]. Available at: https://www.orpha.net/en/disease/detail/106
4. Firth HV, Hurst JA. Autism and autism spectrum disorders. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 350-3.
5. Lord C, Brugha TS, Charman T, Cusack J, Dumas G, Frazier T, et al. Autism spectrum disorder. Nat Rev Dis Primers. 2020;6(1):5. doi:10.1038/s41572-019-0138-4. PMID: 31949163.
6. Tick B, Bolton P, Happé F, Rutter M, Rijsdijk F. Heritability of autism spectrum disorders: a meta-analysis of twin studies. J Child Psychol Psychiatry. 2016;57(5):585-95. doi:10.1111/jcpp.12499. PMID: 26709141.
7. Bai D, Yip BHK, Windham GC, Sourander A, Francis R, Yoffe R, et al. Association of Genetic and Environmental Factors With Autism in a 5-Country Cohort. JAMA Psychiatry. 2019;76(10):1035-1043. doi:10.1001/jamapsychiatry.2019.1411. PMID: 31314057.
8. Sandin S, Lichtenstein P, Kuja-Halkola R, Hultman C, Larsson H, Reichenberg A. The Heritability of Autism Spectrum Disorder. JAMA. 2017;318(12):1182-1184. doi:10.1001/jama.2017.12141. PMID: 28973605.
9. Zeidan J, Fombonne E, Scorah J, Ibrahim A, Durkin MS, Saxena S, et al. Global prevalence of autism: A systematic review update. Autism Res. 2022;15(5):778-790. doi:10.1002/aur.2696. PMID: 35238171.
10. Shaw KA, Williams S, Patrick ME, Valencia-Prado M, Durkin MS, Howerton EM, et al. Prevalence and Early Identification of Autism Spectrum Disorder Among Children Aged 4 and 8 Years - Autism and Developmental Disabilities Monitoring Network, 16 Sites, United States, 2022. MMWR Surveill Summ. 2025;74(2):1-22. doi:10.15585/mmwr.ss7402a1. PMID: 40232988.
11. Oliveira G, Ataíde A, Marques C, Miguel TS, Coutinho AM, Mota-Vieira L, et al. Epidemiology of autism spectrum disorder in Portugal: prevalence, clinical characterization, and medical conditions. Dev Med Child Neurol. 2007;49(10):726-33. doi:10.1111/j.1469-8749.2007.00726.x. PMID: 17880640.
12. Paula CS, Ribeiro SH, Fombonne E, Mercadante MT. Brief report: prevalence of pervasive developmental disorder in Brazil: a pilot study. J Autism Dev Disord. 2011;41(12):1738-42. doi:10.1007/s10803-011-1200-6. PMID: 21337063.
13. Fu JM, Satterstrom FK, Peng M, Brand H, Collins RL, Dong S, et al. Rare coding variation provides insight into the genetic architecture and phenotypic context of autism. Nat Genet. 2022;54(9):1320-1331. doi:10.1038/s41588-022-01104-0. PMID: 35982160.
14. Satterstrom FK, Kosmicki JA, Wang J, Breen MS, De Rubeis S, An JY, et al. Large-Scale Exome Sequencing Study Implicates Both Developmental and Functional Changes in the Neurobiology of Autism. Cell. 2020;180(3):568-584.e23. doi:10.1016/j.cell.2019.12.036. PMID: 31981491.
15. Schaefer GB, Mendelsohn NJ. Clinical genetics evaluation in identifying the etiology of autism spectrum disorders: 2013 guideline revisions. Genet Med. 2013;15(5):399-407. doi:10.1038/gim.2013.32. PMID: 23519317.
16. Phelan K, Rogers RC, Boccuto L. Phelan-McDermid syndrome-SHANK3 related. 2005 May 11 [updated 2024 Jun 6]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1198/
17. Mitchel MW, Myers SM, Heidlebaugh AR, et al. CHD8-related neurodevelopmental disorder with overgrowth. 2022 Oct 27. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK585456/
18. Tramontana T, Yehia L. PTEN hamartoma tumor syndrome. 2001 Nov 29 [updated 2026 Jul 28]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1488/
19. Tammimies K, Marshall CR, Walker S, Kaur G, Thiruvahindrapuram B, Lionel AC, et al. Molecular Diagnostic Yield of Chromosomal Microarray Analysis and Whole-Exome Sequencing in Children With Autism Spectrum Disorder. JAMA. 2015;314(9):895-903. doi:10.1001/jama.2015.10078. PMID: 26325558.
20. Srivastava S, Love-Nichols JA, Dies KA, Ledbetter DH, Martin CL, Chung WK, et al. Meta-analysis and multidisciplinary consensus statement: exome sequencing is a first-tier clinical diagnostic test for individuals with neurodevelopmental disorders. Genet Med. 2019;21(11):2413-2421. doi:10.1038/s41436-019-0554-6. PMID: 31182824.
21. Ozonoff S, Young GS, Bradshaw J, Charman T, Chawarska K, Iverson JM, et al. Familial Recurrence of Autism: Updates From the Baby Siblings Research Consortium. Pediatrics. 2024;154(2):. doi:10.1542/peds.2023-065297. PMID: 39011552.
22. Janssen Pharmaceuticals. RISPERDAL® (risperidone) tablets, oral solution, orally disintegrating tablets — prescribing information [revised May 2026]. Silver Spring (MD): U.S. Food and Drug Administration; 2026. Available at: https://api.fda.gov/drug/label.json?search=openfda.brand_name:%22RISPERDAL%22
23. Otsuka America Pharmaceutical. ABILIFY® (aripiprazole) tablets, oral solution — prescribing information [revised January 2025]. Silver Spring (MD): U.S. Food and Drug Administration; 2025. Available at: https://api.fda.gov/drug/label.json?search=openfda.brand_name:%22ABILIFY%22
24. European Medicines Agency. Slenyto (melatonin) prolonged-release tablets: EPAR — product information (summary of product characteristics), sections 4.1 and 4.2. EMEA/H/C/004425; first authorisation 20 September 2018; extension to neurogenetic disorders (II/0025) 26 August 2024; ADHD indication (II/0028) 17 March 2025; product information updated on 10 June 2026 [accessed 7 November 2026]. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/slenyto
25. Brasil. Ministério da Saúde. Portaria Conjunta SAES/SCTIE n.º 7, de 12 de abril de 2022. Aprova o Protocolo Clínico e Diretrizes Terapêuticas do Comportamento Agressivo no Transtorno do Espectro do Autismo. Brasília: Ministério da Saúde; 2022. Available at: https://www.gov.br/saude/pt-br/assuntos/pcdt/c/comportamento-agressivo-no-transtorno-do-espectro-do-autismo
26. European Medicines Agency. Medicines data file (EPAR): Slenyto (product information updated on 10 June 2026), Abilify — current indication and authorisation dates. Amsterdam: EMA; accessed 4 October 2026. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/slenyto
27. GlaxoSmithKline. WELLCOVORIN (leucovorin calcium) tablets: prescribing information (NDA 018342, efficacy supplement S-015 approved on 9 March 2026; indication 1.2, FOLR1-CFTD; section 14.1). Silver Spring (MD): U.S. Food and Drug Administration; revised March 2026. Available at: https://www.accessdata.fda.gov/drugsatfda_docs/label/2026/018342s015lbl.pdf
28. Haute Autorité de Santé. Trouble du spectre de l'autisme : interventions et parcours de vie du nourrisson, de l'enfant et de l'adolescent [Autism Spectrum Disorder: interventions and life pathways for infants, children and adolescents]. Recommandation de bonne pratique (consensus formalisé), adoptée par le Collège le 8 janvier 2026. Saint-Denis La Plaine: HAS; 12 février 2026. ISBN 978-2-11-179615-7. Available at: https://www.has-sante.fr/jcms/p_4586714
29. Kanner L. Autistic disturbances of affective contact. Nerv Child. 1943;2:217-50. Available at: https://psycnet.apa.org/record/1943-03624-001
30. Asperger H. Die „Autistischen Psychopathen" im Kindesalter. Arch Psychiatr Nervenkr. 1944;117(1):76-136. doi:10.1007/BF01837709.
31. Folstein S, Rutter M. Infantile autism: a genetic study of 21 twin pairs. J Child Psychol Psychiatry. 1977;18(4):297-321. doi:10.1111/j.1469-7610.1977.tb00443.x. PMID: 562353.
32. Iossifov I, O'Roak BJ, Sanders SJ, Ronemus M, Krumm N, Levy D, et al. The contribution of de novo coding mutations to autism spectrum disorder. Nature. 2014;515(7526):216-21. doi:10.1038/nature13908. PMID: 25363768.
33. Gaugler T, Klei L, Sanders SJ, Bodea CA, Goldberg AP, Lee AB, et al. Most genetic risk for autism resides with common variation. Nat Genet. 2014;46(8):881-5. doi:10.1038/ng.3039. PMID: 25038753.
34. Grove J, Ripke S, Als TD, Mattheisen M, Walters RK, Won H, et al. Identification of common genetic risk variants for autism spectrum disorder. Nat Genet. 2019;51(3):431-444. doi:10.1038/s41588-019-0344-8. PMID: 30804558.
35. Kong A, Frigge ML, Masson G, Besenbacher S, Sulem P, Magnusson G, et al. Rate of de novo mutations and the importance of father's age to disease risk. Nature. 2012;488(7412):471-5. doi:10.1038/nature11396. PMID: 22914163.
36. Zhou X, Feliciano P, Shu C, Wang T, Astrovskaya I, Hall JB, et al. Integrating de novo and inherited variants in 42,607 autism cases identifies mutations in new moderate-risk genes. Nat Genet. 2022;54(9):1305-1319. doi:10.1038/s41588-022-01148-2. PMID: 35982159.
37. Simons Foundation Autism Research Initiative. SFARI Gene [database]. New York: Simons Foundation; [accessed 3 October 2026]. Available at: https://gene.sfari.org/
38. Rosenfeld JA, Coe BP, Eichler EE, Cuckle H, Shaffer LG. Estimates of penetrance for recurrent pathogenic copy-number variations. Genet Med. 2013;15(6):478-81. doi:10.1038/gim.2012.164. PMID: 23258348.
39. Jones KL, Jones MC, del Campo M. 22q11.2 microdeletion syndrome (velo-cardio-facial syndrome, DiGeorge syndrome, Shprintzen syndrome). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 386-8.
40. Firth HV, Hurst JA. 47,XXX. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 628.
41. Durand CM, Betancur C, Boeckers TM, Bockmann J, Chaste P, Fauchereau F, et al. Mutations in the gene encoding the synaptic scaffolding protein SHANK3 are associated with autism spectrum disorders. Nat Genet. 2007;39(1):25-7. doi:10.1038/ng1933. PMID: 17173049.
42. Levy T, Foss-Feig JH, Betancur C, Siper PM, Trelles-Thorne MDP, Halpern D, et al. Strong evidence for genotype-phenotype correlations in Phelan-McDermid syndrome: results from the developmental synaptopathies consortium. Hum Mol Genet. 2022;31(4):625-637. doi:10.1093/hmg/ddab280. PMID: 34559195.
43. Bernier R, Golzio C, Xiong B, Stessman HA, Coe BP, Penn O, et al. Disruptive CHD8 mutations define a subtype of autism early in development. Cell. 2014;158(2):263-276. doi:10.1016/j.cell.2014.06.017. PMID: 24998929.
44. Wolff M, Johannesen KM, Hedrich UBS, Masnada S, Rubboli G, Gardella E, et al. Genetic and phenotypic heterogeneity suggest therapeutic implications in SCN2A-related disorders. Brain. 2017;140(5):1316-1336. doi:10.1093/brain/awx054. PMID: 28379373.
45. Firth HV, Hurst JA. Macrocephaly. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 224-6.
46. Christensen J, Grønborg TK, Sørensen MJ, Schendel D, Parner ET, Pedersen LH, et al. Prenatal valproate exposure and risk of autism spectrum disorders and childhood autism. JAMA. 2013;309(16):1696-703. doi:10.1001/jama.2013.2270. PMID: 23613074.
47. Correia LP, Almeida TP. Síndrome decorrente do uso da talidomida no pré-natal. In: Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 187.
48. Jones KL, Jones MC, del Campo M. Hyperthermia-induced spectrum of defects. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 808.
49. U.S. Food and Drug Administration. FDA responds to evidence of possible association between autism and acetaminophen use during pregnancy [press release]. Silver Spring (MD): FDA; 2025 Sep 22. Available at: https://content.govdelivery.com/accounts/USFDA/bulletins/3f3def0
50. European Medicines Agency. Use of paracetamol during pregnancy unchanged in the EU [press release]. Amsterdam: EMA; 2025 Sep 23. Available at: https://www.ema.europa.eu/en/news/use-paracetamol-during-pregnancy-unchanged-eu
51. Ahlqvist VH, Sjöqvist H, Dalman C, Karlsson H, Stephansson O, Johansson S, et al. Acetaminophen use during pregnancy and children's risk of autism, ADHD, and intellectual disability. JAMA. 2024;331(14):1205-14. doi:10.1001/jama.2024.3172. PMID: 38592388.
52. Hirota T, King BH. Autism Spectrum Disorder: A Review. JAMA. 2023;329(2):157-168. doi:10.1001/jama.2022.23661. PMID: 36625807.
53. Hyman SL, Levy SE, Myers SM. Identification, Evaluation, and Management of Children With Autism Spectrum Disorder. Pediatrics. 2020;145(1):. doi:10.1542/peds.2019-3447. PMID: 31843864.
54. Srivastava S, Sahin M, Buxbaum JD, Berry-Kravis E, Soorya LV, Thurm A, et al. Updated consensus guidelines on the management of Phelan-McDermid syndrome. Am J Med Genet A. 2023;191(8):2015-2044. doi:10.1002/ajmg.a.63312. PMID: 37392087.
55. Dhawan A, Liu D, Baitamouni S, Anthony K, Srivastava S, Hardan AY, et al. Neurodevelopmental and Neurologic Manifestations of PTEN Hamartoma Tumor Syndrome: Management Recommendations. Neurol Genet. 2025;11(5):e200299. doi:10.1212/NXG.0000000000200299. PMID: 41080814.
56. Jones KL, Jones MC, del Campo M. Bannayan-Riley-Ruvalcaba syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 742-3.
57. Yehia L, Plitt G, Tushar AM, Joo J, Burke CA, Campbell SC, et al. Longitudinal Analysis of Cancer Risk in Children and Adults With Germline PTEN Variants. JAMA Netw Open. 2023;6(4):e239705. doi:10.1001/jamanetworkopen.2023.9705. PMID: 37093598.
58. Firth HV, Hurst JA. Cowden syndrome (PTEN hamartoma tumour syndrome (PHTS)). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 564-7.
59. Manickam K, McClain MR, Demmer LA, Biswas S, Kearney HM, Malinowski J, et al. Exome and genome sequencing for pediatric patients with congenital anomalies or intellectual disability: an evidence-based clinical guideline of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(11):2029-2037. doi:10.1038/s41436-021-01242-6. PMID: 34211152.
60. Miller DT, Adam MP, Aradhya S, Biesecker LG, Brothman AR, Carter NP, et al. Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies. Am J Hum Genet. 2010;86(5):749-64. doi:10.1016/j.ajhg.2010.04.006. PMID: 20466091.
61. Butler MG, Dasouki MJ, Zhou XP, Talebizadeh Z, Brown M, Takahashi TN, et al. Subset of individuals with autism spectrum disorders and extreme macrocephaly associated with germline PTEN tumour suppressor gene mutations. J Med Genet. 2005;42(4):318-21. doi:10.1136/jmg.2004.024646. PMID: 15805158.
62. Tatton-Brown K, Loveday C, Yost S, Clarke M, Ramsay E, Zachariou A, et al. Mutations in Epigenetic Regulation Genes Are a Major Cause of Overgrowth with Intellectual Disability. Am J Hum Genet. 2017;100(5):725-736. doi:10.1016/j.ajhg.2017.03.010. PMID: 28475857.
63. Brunklaus A, Feng T, Brünger T, Perez-Palma E, Heyne H, Matthews E, et al. Gene variant effects across sodium channelopathies predict function and guide precision therapy. Brain. 2022;145(12):4275-86. doi:10.1093/brain/awac006. PMID: 35037686.
64. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-24. doi:10.1038/gim.2015.30. PMID: 25741868.
65. Riggs ER, Andersen EF, Cherry AM, Kantarci S, Kearney H, Patel A, et al. Technical standards for the interpretation and reporting of constitutional copy-number variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics (ACMG) and the Clinical Genome Resource (ClinGen). Genet Med. 2020;22(2):245-57. doi:10.1038/s41436-019-0686-8. PMID: 31690835.
66. Morales J, Pujar S, Loveland JE, Astashyn A, Bennett R, Berry A, et al. A joint NCBI and EMBL-EBI transcript set for clinical genomics and research. Nature. 2022;604(7905):310-5. doi:10.1038/s41586-022-04558-8. PMID: 35388217. MANE file GRCh38 v1.5: https://ftp.ncbi.nlm.nih.gov/refseq/MANE/
67. Jones KL, Jones MC, del Campo M. Xq distal duplication or disomy (Xq27-q28 terminal duplications including MECP2 duplication). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 96-7.
68. Jones KL, Jones MC, del Campo M. Kabuki syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 138.
69. Niinikoski H, Schiff M, Tanner L. Disorders of Amino Acid Transport at the Cell Membrane. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 25, section 25.6 SLC7A5/Brain Neutral Amino Acid Transporter Deficiency. doi:10.1007/978-3-662-63123-2_25
70. García-Cazorla Á, Artuch R, Pearl PL. Disorders of Neurotransmission. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 30, section 30.6 Synaptic Vesicle Disorders. doi:10.1007/978-3-662-63123-2_30
71. García-Cazorla Á, Dionisi-Vici C, Saudubray JM. Disorders of Cellular Trafficking. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 44, section 44.3.1 Neurological Manifestations. doi:10.1007/978-3-662-63123-2_44
72. National Institute for Health and Care Excellence. Autism spectrum disorder in under 19s: support and management. Clinical guideline CG170. London: NICE; 2013 [updated 2021]. Available at: https://www.nice.org.uk/guidance/cg170
73. Dawson G, Rogers S, Munson J, Smith M, Winter J, Greenson J, et al. Randomized, controlled trial of an intervention for toddlers with autism: the Early Start Denver Model. Pediatrics. 2010;125(1):e17-23. doi:10.1542/peds.2009-0958. PMID: 19948568.
74. Pickles A, Le Couteur A, Leadbitter K, Salomone E, Cole-Fletcher R, Tobin H, et al. Parent-mediated social communication therapy for young children with autism (PACT): long-term follow-up of a randomised controlled trial. Lancet. 2016;388(10059):2501-2509. doi:10.1016/S0140-6736(16)31229-6. PMID: 27793431.
75. McCracken JT, McGough J, Shah B, Cronin P, Hong D, Aman MG, et al. Risperidone in children with autism and serious behavioral problems. N Engl J Med. 2002;347(5):314-21. doi:10.1056/NEJMoa013171. PMID: 12151468.
76. Owen R, Sikich L, Marcus RN, Corey-Lisle P, Manos G, McQuade RD, et al. Aripiprazole in the treatment of irritability in children and adolescents with autistic disorder. Pediatrics. 2009;124(6):1533-40. doi:10.1542/peds.2008-3782. PMID: 19948625.
77. U.S. Food and Drug Administration. Drugs@FDA: NDA 020272 (Risperdal), efficacy supplement approved on 2006-10-06; NDA 021436 (Abilify), efficacy supplement approved on 2009-11-19. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/
78. European Medicines Agency. Abilify (aripiprazole): EPAR — medicine overview. Amsterdam: EMA; [accessed 3 October 2026]. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/abilify
79. European Medicines Agency. Risperdal and associated names: Article 30 referral (CHMP opinion of 24 July 2008; European Commission decision of 7 October 2008) and harmonised risperidone SmPC, section 4.1. Amsterdam: EMA; 2008. Available at: https://www.ema.europa.eu/en/medicines/human/referrals/risperdal ; text of section 4.1 at: https://www.medicines.org.uk/emc/product/3957/smpc
80. European Medicines Agency. Risperdal — Article 30 referral (CHMP opinion of 24 July 2008; European Commission decision of 7 October 2008). Amsterdam: EMA; 2008. Available at: https://www.ema.europa.eu/en/medicines/human/referrals/risperdal
81. Gringras P, Nir T, Breddy J, Frydman-Marom A, Findling RL. Efficacy and Safety of Pediatric Prolonged-Release Melatonin for Insomnia in Children With Autism Spectrum Disorder. J Am Acad Child Adolesc Psychiatry. 2017;56(11):948-957.e4. doi:10.1016/j.jaac.2017.09.414. PMID: 29096777.
82. ClinGen Actionability Working Group. PTEN hamartoma tumor syndrome — Cowden syndrome (report AC025), adult and paediatric contexts. Clinical Genome Resource; [accessed 3 October 2026]. Available at: https://actionability.clinicalgenome.org/ac/
83. Sikich L, Kolevzon A, King BH, McDougle CJ, Sanders KB, Kim SJ, et al. Intranasal oxytocin in children and adolescents with autism spectrum disorder. N Engl J Med. 2021;385(16):1462-73. doi:10.1056/NEJMoa2103583. PMID: 34644471.
84. Jacob S, Veenstra-VanderWeele J, Murphy D, McCracken J, Smith J, Sanders K, et al. Efficacy and safety of balovaptan for socialisation and communication difficulties in autistic adults in North America and Europe: a phase 3, randomised, placebo-controlled trial. Lancet Psychiatry. 2022;9(3):199-210. doi:10.1016/S2215-0366(21)00429-6. PMID: 35151410.
85. Hollander E, Jacob S, Jou R, McNamara N, Sikich L, Tobe R, et al. Balovaptan vs placebo for social communication in childhood autism spectrum disorder: a randomized clinical trial. JAMA Psychiatry. 2022;79(8):760-9. doi:10.1001/jamapsychiatry.2022.1717. PMID: 35793101.
86. Fuentes J, Parellada M, Georgoula C, Oliveira G, Marret S, Crutel V, et al. Bumetanide oral solution for the treatment of children and adolescents with autism spectrum disorder: results from two randomized phase III studies. Autism Res. 2023;16(10):2021-34. doi:10.1002/aur.3005. PMID: 37794745.
87. U.S. Food and Drug Administration. FDA takes action to make a treatment available for autism symptoms [press release]. Silver Spring (MD): FDA; 2025 Sep 22. Available at: https://content.govdelivery.com/accounts/USFDA/bulletins/3f3deae
88. Panda PK, Sharawat IK, Saha S, Gupta D, Palayullakandi A, Meena K. Retraction Note: Efficacy of oral folinic acid supplementation in children with autism spectrum disorder: a randomized double-blind, placebo-controlled trial. Eur J Pediatr. 2026;185(2):109. doi:10.1007/s00431-026-06769-x. PMID: 41606385. (Retracted article: Eur J Pediatr. 2024;183(11):4827-35. PMID: 39243316.)
89. American Academy of Pediatrics. Use of leucovorin in autistic pediatric patients: frequently asked questions (FAQs) for pediatricians and other prescribing pediatric clinicians. Itasca (IL): AAP; accessed 4 October 2026. Available at: https://www.aap.org/en/patient-care/autism/use-of-leucovorin-in-autistic-pediatric-patients/frequently-asked-questions-faqs-for-pediatricians-and-other-prescribing-pediatric-clinicians/
90. Rome BN, Kim J, Figueroa JF, Kesselheim AS. Use of leucovorin after the FDA announcement of its use for autism spectrum disorders. Pediatrics. 2026;158(2):e2026076794. doi:10.1542/peds.2026-076794. PMID: 42260633.
91. Marcus RN, Owen R, Kamen L, Manos G, McQuade RD, Carson WH, et al. A placebo-controlled, fixed-dose study of aripiprazole in children and adolescents with irritability associated with autistic disorder. J Am Acad Child Adolesc Psychiatry. 2009;48(11):1110-9. doi:10.1097/CHI.0b013e3181b76658. PMID: 19797985.
92. Ozonoff S, Young GS, Carter A, Messinger D, Yirmiya N, Zwaigenbaum L, et al. Recurrence risk for autism spectrum disorders: a Baby Siblings Research Consortium study. Pediatrics. 2011;128(3):e488-95. doi:10.1542/peds.2010-2825. PMID: 21844053.
93. Sandin S, Lichtenstein P, Kuja-Halkola R, Larsson H, Hultman CM, Reichenberg A. The familial risk of autism. JAMA. 2014;311(17):1770-7. doi:10.1001/jama.2014.4144. PMID: 24794370.
94. Autism-Europe. Brussels. Available at: https://www.autismeurope.org ; Federação Portuguesa de Autismo: https://www.fpda.pt
95. European Reference Network on Rare Malformation Syndromes, Intellectual and Other Neurodevelopmental Disorders (ERN ITHACA). Available at: https://ern-ithaca.eu
96. European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). Available at: https://www.genturis.eu
97. White SL, Jamil T, Bell C, Fishbein L, Haugen BR, Gignoux CR, et al. Population prevalence of the major thyroid cancer-associated syndromes. J Clin Endocrinol Metab. 2025;110(12):e4049-54. doi:10.1210/clinem/dgaf236. PMID: 40231587.
98. Moreno Fernandes M, Rodrigues Neto M, Sá Pinto M, Pena Fernandes T, Maia AC, Ayres Pereira I, et al. A Novel PTEN Frameshift Variant in a Child With Autism Spectrum Disorder and Macrocephaly: A Case Report. Cureus. 2026;18(1):e100834. doi:10.7759/cureus.100834. PMID: 41646565.