ZMIZ1-Related Neurodevelopmental Disorder

Neurodevelopmental disorder with dysmorphic facies and distal skeletal anomalies (NEDDFSA)

Literature and regulatory information updated: 28 October 2026.

0. Initial Reference Table

1. Introduction

1.1 Definition and scope

ZMIZ1-related neurodevelopmental disorder is an autosomal dominant genetic disease caused by heterozygous pathogenic variants in the ZMIZ1 gene or by chromosomal rearrangements that disrupt the gene or its regulatory region. OMIM classifies it as entry #618659, under the designation neurodevelopmental disorder with dysmorphic facies and distal skeletal anomalies (NEDDFSA), and describes it as a global neurodevelopmental disorder with highly variable features: feeding difficulties, poor growth and hypotonia from the first year of life, followed by mild motor delay, poor language acquisition and behavioural abnormalities; intellectual development ranges from severe ID without speech to mild ID compatible with attendance at special education; facial dysmorphism with ocular anomalies, joint hypermobility and mild skeletal anomalies of the hands and feet complete the picture [1]. This article covers a single entity, without numbered subtypes. Orphanet does not assign it a code of its own and includes the gene in group ORPHA:528084 of non-specific syndromic ID, defined by the association of ID with other variable anomalies in the absence of a clinically recognisable syndrome [14].

1.2 From gene to disease

The gene was isolated in 1999 under the designation KIAA1224, from a human fetal brain complementary DNA (cDNA) library, by Nagase and colleagues, and later received the names ZIMP10 and RAI17 before the current symbol approved by the HGNC (HUGO Gene Nomenclature Committee) [22, 23]. The first link to a neurodevelopmental disease emerged in 2015, when Córdova-Fletes and colleagues described a Mexican girl with ID and neuropsychiatric abnormalities carrying a de novo balanced translocation t(10;19)(q22.3;q13.33); fine mapping of the breakpoints by array painting showed disruption of ZMIZ1 on chromosome 10 and of PRR12 on chromosome 19, with reciprocal fusion transcripts that introduced premature termination codons [24]. That study could not attribute the phenotype to either of the two genes, since both are expressed in the brain and both could be haploinsufficient [11, 24]. Also in 2019, a French low-coverage genome sequencing study of 55 patients with balanced rearrangements and developmental disorders identified ZMIZ1 among the disrupted genes and among the genes affected by a position effect [19].

1.3 The founding series and delineation of the syndrome

The disease was delineated by Carapito and colleagues in the American Journal of Human Genetics, published online on 10 January 2019 [11]. The series brought together 19 individuals (16 unrelated families) through a transatlantic collaboration of French (Strasbourg, Nantes, Dijon, Rouen, Paris and Lyon), North American, British and Polish centres, using the case-matching platforms GeneMatcher, MyGene2, DECIPHER and Matchmaker Exchange [11]. Fourteen unrelated individuals had de novo point variants or small insertions and deletions, three siblings shared a single-nucleotide insertion (parents unavailable for study) and two had de novo balanced translocations; the 13 distinct point variants were absent from gnomAD (Genome Aggregation Database), the Exome Variant Server and an in-house database of 550 exomes [11]. An erratum published in 2020 corrected the supplementary table: subjects 13 to 15 carry the same variant (and not different variants) and subjects 18 and 19 have translocations, not duplications [25]. OMIM entry #618659, created in November 2019, summarises this series and established the designation NEDDFSA; its text counts 17 patients with point variants (14 unrelated and the three adopted siblings), without the two translocation carriers, and the gene entry (*607159) selects six alleles from the founding series (607159.0001 to 607159.0006) [1, 6]. According to OMIM, only two patients in the series had severe ID, among them a 19-year-old woman without speech, and several attended special education [1].

1.4 Subsequent developments

In 2020, Latchman and colleagues demonstrated autosomal dominant transmission in a family from Miami (father and two affected children) and showed that severity varies within the same family, which supports testing the parents in search of a mildly affected parent [12]. Between 2021 and 2026, isolated cases were published from Thailand, China, Canada and Sweden, together with an ophthalmological review, a prenatal diagnosis and the report of a woman followed up to the age of 52, the oldest adult patient described [3, 17, 22, 23, 26–29]. In 2022, the ClinGen Intellectual Disability and Autism expert panel classified the relationship between ZMIZ1 and complex neurodevelopmental disorder as definitive [15]. In 2025, Cormier and colleagues gathered 36 individuals from 15 studies and analysed the association between the affected protein domain and the phenotype [4]; in 2026, the same group published a mouse model with forebrain deletion of Zmiz1, showing cortical microcephaly, corpus callosum dysgenesis and behavioural changes similar to those associated with autism [30].

1.5 Nomenclature

The literature uses different designations for the same entity. OMIM and most clinical reports adopt NEDDFSA; ClinVar uses ZMIZ1-related disorder; the most recent reports refer to ZMIZ1-associated neurodevelopmental disorder; the Galazo group uses ZMIZ1 syndromic neurodevelopmental disorder; ClinGen classifies the gene–disease relationship under the generic MONDO term complex neurodevelopmental disorder [1–4, 15]. The designation “distal skeletal anomalies” refers to mild abnormalities of the fingers and toes (brachydactyly, cutaneous syndactyly, cone-shaped epiphyses, clubfoot) and not to a generalised skeletal dysplasia [11].

2. Molecular Genetics and Pathophysiology

2.1 The gene and the protein

ZMIZ1 is located at 10q22.3 (GRCh38 coordinates 10:79 068 966–79 316 519) and encodes a protein of 1067 amino acids (NP_065071.1), with a calculated mass of 123 kDa [6, 22]. The first mapping, by genomic sequence analysis in 2003, placed the gene in the interval 10q22.1–q22.3, and the same work described 21 exons, a number repeated by the 2022 Chinese report [6, 22]; the canonical transcript ENST00000334512 (ZMIZ1-201) has, however, 25 exons in Ensembl (release 116), a count that includes the non-coding exons and that is consistent with the location of the c.3097-2A>G variant in intron 24, the last intron of the gene [6, 10]. The full-length protein, isolated by 5′ extension (RACE) from brain and prostate libraries, contains the MSX-interacting zinc finger domain (MIZ domain), a nuclear localisation signal and two proline-rich regions, and migrates at about 130 kDa after in vitro translation [6]. The MANE Select transcript is NM_020338.4, corresponding to ENST00000334512 [8, 9]. The founding series numbered the variants on NM_020338.3; the c. positions of the published variants are identical in versions .3 and .4 [8, 11]. The protein belongs to the family of PIAS-like (protein inhibitor of activated STAT) transcriptional coregulators: it does not bind DNA directly and modulates transcription factors, usually through SUMOylation, the covalent attachment of small ubiquitin-like modifiers (SUMO) [5, 11]. The domain organisation is summarised in the table, with the coordinates from the founding series and, where they differ, those obtained by motif search in the report by Lu and colleagues (see figure 2 of Carapito and colleagues, https://pmc.ncbi.nlm.nih.gov/articles/PMC6369415/figure/fig2/) [11, 22].

Table sources: coordinates and functions [11, 22]; variant type by domain in the pooled analysis of 36 individuals [4].

2.2 Function of ZMIZ1

ZMIZ1 enhances the transcriptional activity of p53 and of the androgen receptor, colocalises with this receptor and with SUMO-1 in the nucleus and promotes its SUMOylation; it coactivates the SMAD3/SMAD4 complex in transforming growth factor beta (TGF-β) signalling; and it activates a subset of NOTCH1 target genes, including MYC, through direct interaction of the TPR domain with NOTCH1, which explains its role in T-lymphocyte development [5, 11, 20]. In the studies that characterised it as an androgen receptor coregulator, binding occurs between the central region of ZMIZ1 and the transactivation domain of the receptor, the C-terminal proline-rich region has intrinsic transactivation activity, and the protein accompanies the receptor and SUMO-1 at replication foci throughout S phase; enhancement of receptor activity depends on receptor SUMOylation, because it disappears with receptor mutants lacking SUMOylation sites [6]. Conditional deletion of Zmiz1 in the mouse blocked T-lymphocyte development and caused transient intestinal goblet cell hyperplasia, as does deletion of Notch1, but, unlike the latter, it did not alter intestinal homeostasis or myeloid suppression; the gene was required to initiate and maintain NOTCH1-induced T-cell acute lymphoblastic leukaemia, and binding occurs between the TPR domain of ZMIZ1 and the RAM domain of NOTCH1, with selective regulation of a subset of NOTCH1 targets, above all MYC [6]. In the nervous system, the protein interacts with the neuronal chromatin remodelling complex nBAF (including BRG1, encoded by SMARCA4) and with the AP-1 complex (activator protein 1), which regulate genes involved in synapse and dendrite formation [11, 24]. In the human adult, ZMIZ1 RNA was detected by RT-PCR in all tissues and brain regions studied, with the highest levels in the heart, the ovary and the brain and, within the brain, in the amygdala; by Northern blot, the 7.2 kb transcript was most abundant in the ovary, lower in the prostate, spleen and testis, and almost absent in the thymus, intestine, colon and peripheral blood leucocytes [6]. In the mouse embryo, Zmiz1 and its paralogue Zmiz2 are expressed in partially overlapping patterns from embryonic day 7.5 to mid-gestation, predominantly in neural tissues at the early stages, with a fall in expression at day 12.5 [6]. ZMIZ1 is highly expressed in the human and murine embryonic brain, with enrichment in the cortex, hippocampus and cerebellum, and in craniofacial tissue; in the limbs of the mouse embryo, expression is restricted to the interdigital region at the stage corresponding to embryonic day 12, which is consistent with the syndactyly observed in 6 of the 19 patients of the founding series [11, 31]. Homozygous null mice die during organogenesis owing to failure of yolk sac vascular remodelling [11].

2.3 Genetic constraint and gene–disease validity

In gnomAD (version 4), ZMIZ1 has 17 observed loss-of-function variants for 113.9 expected, an observed/expected ratio of 0.149 (90% confidence interval 0.102–0.224, whose rounded upper bound, the LOEUF, loss-of-function observed/expected upper bound fraction, is 0.22) and a probability of loss-of-function intolerance (pLI) of 1.00; for missense variants, the ratio is 0.717 and the constraint Z score is 4.65 [9]. These values indicate intolerance to the loss of one allele and moderate constraint on amino acid substitutions. The DECIPHER haploinsufficiency probability score (%HI) is 11.32 [16]. Gene–disease validity is definitive in ClinGen, and SFARI Gene includes the gene in category 2S, with 15 reports (8 of syndromic autism) and 26 rare variants [7, 15].

2.4 Disease mechanism

The molecular mechanism is unresolved, and the sources point to different hypotheses depending on the variant type. In favour of haploinsufficiency are the loss-of-function intolerance in gnomAD, the rearrangements that disrupt the gene or an upstream enhancer, the ZMIZ1/PRR12 fusion with premature termination codons and the statement, in a 2026 review, that reduction to about half the dose of ZMIZ1 is no longer tolerated by the developing nervous system [9, 11, 20, 24]. In the founding series, however, ZMIZ1 messenger RNA (mRNA) expression in the blood of five patients with point variants was similar to that of controls, the mutant proteins p.Lys91Arg, p.Thr300Met and p.Thr1038fs were stable and localised to the nucleus, and the last-exon frameshift variants (p.Thr1038AsnfsTer4, p.Phe1008LeufsTer7) are expected to escape nonsense-mediated mRNA decay (NMD), giving rise to proteins lacking the TAD [11]. ZMIZ1 expression was reduced only in the lymphoblastoid line of the patient with the t(X;10) translocation, which disrupts an enhancer located 276 kb upstream of the gene [11].

The functional assays of the founding series showed that three mutants reduced androgen receptor coactivation in a luciferase assay in HEK293T cells (without statistical significance for p.Lys91Arg) and that overexpression of human mutants in the mouse cortex by in utero electroporation (embryonic day 14.5 to 18.5) retained about 60% of neurons in the ventricular, subventricular and intermediate zones for p.Lys91Arg and p.Thr300Met, and about 40% for p.Thr1038fs, with rounded cells or cells with misoriented processes, that is, a defect of migration and polarisation of pyramidal neurons; the wild-type allele only partially rescued the p.Thr300Met defect (figure 3, https://pmc.ncbi.nlm.nih.gov/articles/PMC6369415/figure/fig3/) [11]. Because a defect obtained by overexpression suggests a dominant effect rather than simple loss of one allele, the authors concluded that further studies were needed to define the mechanism [11].

Subsequent studies of variants in the alanine-rich region reinforce this heterogeneity. The in-frame deletion p.(Val288_Ala293del) reduced expression of the mutant protein on Western blot and immunofluorescence [27], whereas the substitution p.(Ala304Pro) increased ZMIZ1 mRNA and protein, altered its subcellular localisation and, in human skeletal myoblasts, increased proliferation, migration and invasion, with dysregulation of TGFβ1, SMAD2, SMAD3, SMAD4, CTGF and ERK2 and a new interaction with GTF2I identified by co-immunoprecipitation and mass spectrometry; the authors acknowledge the limitations of a single proband and of an overexpression system [23]. GTF2I is one of the genes in the 7q11.23 region deleted in Williams-Beuren syndrome, and a 2026 commentary proposed this interaction as a possible contributor to the neurological phenotype [32]. Structural modelling with AlphaFold suggests that missense variants in the alanine-rich and TPR domains alter the relative position of the domains, a prediction without experimental validation [4].

2.5 Variant spectrum

Causal variants include missense variants clustered in the alanine-rich region, frameshift and nonsense variants scattered throughout the gene, concentrated in the proline-rich domain and the TAD, splice variants, in-frame deletions, and balanced chromosomal rearrangements [4, 11]. In ClinVar (search of 28 October 2026), ZMIZ1 had 914 records, of which 37 were classified as pathogenic, 47 as likely pathogenic and 414 as of uncertain significance (VUS); among pathogenic or likely pathogenic records, frameshift, nonsense and splice variants predominate across the whole gene and, in the alanine-rich region, missense variants and in-frame duplications or deletions (c.858_875dup, c.858_878dup, c.858_878del, c.845_862dup, c.857_872del, c.863T>G p.(Val288Gly)); a genomic deletion NC_000010.11:g.79055368_79303220del classified as pathogenic is also listed [2]. The table brings together the published variants with the phenotype described, normalised in VariantValidator to NM_020338.4 (GRCh38) [8].

Three published notations do not correspond to the normalised form and should not be reproduced: the c.2752dupC variant of the founding series is written c.2758dup, by application of the HGVS (Human Genome Variation Society) 3′ alignment rule; the Thai splice variant appears in the report by Lu and colleagues as c.1497+2T>C, whereas the correct form is c.1491+2T>C; and the variant of the Chinese boy reported in 2026 is c.910G>C, although the discussion of the article also writes it as c.910G>A [8, 22, 23]. In the case table of Li and colleagues, subject 16 of the founding series appears with p.Thr296Lys, but the original variant table and ClinVar give c.887C>T p.(Thr296Ile) [11, 23]. In the founding series, the genomic description of c.2835delT indicated the loss of two nucleotides, which does not match the normalised single-nucleotide deletion [8, 11].

Three rearrangements have been described: the translocation t(10;19)(q22.3;q13.33) with reciprocal ZMIZ1/PRR12 fusion [24]; the translocation 46,XX,t(X;10)(q27.3;q22.3), with a breakpoint 276 kb upstream of ZMIZ1, without disruption of the gene but with disruption of an enhancer (chr10:80559750–80568675, GRCh37) and reduced ZMIZ1 expression; and the translocation 46,XX,t(10;12)(q22.2;q24.3), with a breakpoint within ZMIZ1 [11]. No founder variants or population-specific recurrent variants have been described; the only variant repeated in unrelated patients is c.899C>T p.(Thr300Met) [3, 11].

2.6 Genotype–phenotype correlations

The pooled analysis of 36 individuals with coding variants, which excludes translocations and splice variants, distributed the variants by domain and compared phenotypes [4]. The eight variants in the alanine-rich region were all missense or in-frame deletions and all carriers had a neurodevelopmental disorder, with isolated ID in 5/8 (62.5%), ID with autism spectrum disorder (ASD) in 1/8 and isolated ASD in 1/8; in the TPR domain (n=5), ID in 2/5, ID with ASD in 2/5 and ASD without ID in 1/5; in the central proline-rich domain (n=10, 9 frameshift), ID with ASD in 4/10, ID with other manifestations in 3/10 and isolated ASD in 3/10; in the SP-RING/MIZ domain (n=6, all missense), isolated ASD in 4/6; in the TAD (n=6, 5 frameshift), ASD with ID in 3/6 and ID with other manifestations in 3/6 [4]. In the subset of 21 individuals with detailed phenotyping, ASD was less frequent with variants in the alanine-rich region (14%) than in the TPR (60%) and proline-rich (66%) domains, whereas motor delay was more frequent in the alanine-rich region (71%, versus 44% and 40%); facial dysmorphism and distal skeletal anomalies were present in all carriers of variants in the alanine-rich region [4]. Abnormalities of brain and cranial morphology were frequent with any domain, with no demonstrable association [4]. These results derive from small denominators and heterogeneous clinical descriptions, and the article itself describes the 35 variants in an internally inconsistent way (at one point as 2 synonymous and 33 missense, at another including frameshift and nonsense variants); they have the value of a hypothesis and not of a prognostic rule. A review of 25 cases and one fetus found no clear genotype–phenotype correlation [17].

2.7 Penetrance and expressivity

Expressivity is highly variable between families and within the same family: in the family of Latchman and colleagues, the father and children had short stature and ID, but the severity of the ID and the palpebral ptosis varied and only one child had sensorineural hearing loss [12]. Penetrance was considered almost complete in a review released as a preprint [33], but the French network for molecular diagnosis of ID identified, between 2020 and 2022, patients with pathogenic or likely pathogenic variants inherited from an asymptomatic parent in 11 genes usually associated with de novo variants, including ZMIZ1; in the six families in which the grandparents were studied, the variant had arisen de novo in the healthy parent [13]. The Thai case illustrates the difficulty: the c.1491+2T>C variant, which causes in-frame skipping of exon 14 (26 amino acids of the proline-rich region, with preservation of the TAD and the NLS) demonstrated in blood mRNA, was inherited from an asymptomatic, university-educated father, and the daughter's phenotype was better explained by biallelic OTUD6B variants; the authors considered that the ZMIZ1 variant might be non-pathogenic or pathogenic with incomplete penetrance [26]. Splice variants at the same position exist in gnomAD (c.1491+1G>A in one individual and c.1491+2T>A in five), which reinforces caution [26].

2.8 Animal models

Mice with forebrain-restricted deletion of Zmiz1 (Emx1-Cre;Zmiz1 fl/fl) have cortical microcephaly (reduced cortical area from postnatal day 7), a thinner cortex with disproportionate reduction of layer 6, fewer progenitors and neurons per cortical column, corpus callosum dysgenesis and abnormal differentiation of the callosal neurons of layers 2/3, with alterations in motor activity, anxiety, communication, social interaction and sensory gating, marked by sex differences [30]. Combined RNA sequencing and chromatin immunoprecipitation identified Lhx2, Auts2, EfnB2, EphA5 and Unc5d as direct targets; recombinant ephrin-B2 rescued the neurite branching deficit in culture [30]. These data, discussed in an editorial that highlights ZMIZ1 as a risk gene for neurodevelopmental disorders, are preclinical [34].

2.9 Other associations of the gene

ZMIZ1 has roles outside Mendelian disease, with no demonstrated relationship to NEDDFSA: the somatic ZMIZ1-ABL1 fusion has been described in leukaemias; common variants and expression quantitative trait loci (eQTL) are associated with autoimmune diseases (multiple sclerosis, coeliac disease, vitiligo, Crohn's disease), type 2 diabetes and cancers; and ZMIZ1 cooperates with NOTCH1 in the activation of MYC in T-cell acute lymphoblastic leukaemia [20]. The sources consulted do not describe allelic Mendelian diseases other than NEDDFSA.

3. Clinical Spectrum and Manifestations by System

3.1 Nature of the evidence

There is no prospective natural history study. Clinical knowledge comes from a founding series of 19 individuals aged 1 to 20 years at last assessment (11 female and 8 male) [11], from case and family reports published between 2020 and 2026 and from reviews that pool these reports with variable denominators [4, 17, 23, 26, 28]. Because the reports describe children already in rehabilitation and rarely state the interventions, it is not possible to separate spontaneous course from course under supportive treatment. The following frequencies should be read as proportions in series with selection bias towards syndromic cases referred to genetics, and not as individual risks.

3.2 Frequencies in the founding series

Source: table 1 and text of Carapito and colleagues; counts of positive over assessed [11]. The facial and limb photographs are in figure 1 of that article (https://pmc.ncbi.nlm.nih.gov/articles/PMC6369415/figure/fig1/).

3.3 Neurodevelopment, cognition and behaviour

All patients in the founding series had ID or developmental delay, and severity ranges from severe ID without speech to mild ID with schooling in special education [1, 11]. Language is the most affected area (15/17), followed by behaviour (13/16) and motor skills (12/16) [11]. In the Chinese case assessed at 5 years and 1 month with the WPPSI-IV (Wechsler Preschool and Primary Scale of Intelligence, 4th edition), the full-scale intelligence quotient (IQ) was 75, at the threshold between borderline functioning and mild ID [22], and the 52-year-old adult patient had a low-average IQ [3], which shows that cognitive functioning may lie above the ID range. ASD, attention deficit hyperactivity disorder (ADHD), aggressiveness and autistic behaviour are part of the spectrum annotated in OMIM and in the reviews [1, 20]. In the subset of 21 individuals in the pooled analysis, ASD was present in 14% of carriers of variants in the alanine-rich region and in about two thirds of carriers of variants in the TPR and proline-rich domains; ADHD and epileptic seizures occurred in fewer than 30% in all groups, emotional dysregulation in 40% in the TPR domain and sleep disturbances in 14% [4]. ZMIZ1 has also been identified in large sequencing cohorts of patients with autism, which justifies its inclusion in ASD panels [4, 20].

3.4 Nervous system and neuroimaging

Hypotonia (10/16) is frequent in infancy and epileptic seizures are infrequent (3/18) [11]. In a compilation of 23 patients from the literature, epilepsy was recorded in 3/18 assessed (16.7%), hypotonia in 12/19 (63.2%) and central nervous system anomalies in 7/17 (41.2%) [26]. The 2025 pooled analysis reviewed the MRI scans of 14 individuals: microcephaly or abnormal skull shape in 9/14 (64%); abnormalities of the cortex, midline or ventricles in 8/14 (57%), including white matter and periventricular heterotopias and abnormalities of the septum pellucidum and the ventricles; thinner cortex or white matter and abnormal gyration in 7/14 (50%); corpus callosum abnormalities with heterotopias in 3/14 (21%); enlarged meningeal space in 2/14 (14%); and cerebellar atrophy in 1/14 (7%) [4]. Taking MRI, computed tomography and cranial measurements together, 78% of those assessed had abnormalities [4]. The phenotype annotated in OMIM also includes delayed myelination, hypoplasia of the corpus callosum, periventricular nodular heterotopia, ventriculomegaly and cerebral and cerebellar atrophy [1]. In OMIM's description of the founding series, brain imaging, when performed, was normal in some patients and in the others showed non-specific abnormalities (enlarged ventricles, cerebral atrophy, delayed myelination, thin corpus callosum, cerebellar atrophy), with nodular heterotopia in two [1]. The cortical neuronal heterotopias found at autopsy in the 52-year-old patient are consistent with the neuronal migration defect observed in the in utero electroporation model [3, 11].

3.5 Growth and feeding

Growth delay (10/17) and feeding difficulties (9/17) begin in infancy [11]. Of the 13 patients with known birth weight, 4 had a weight ≤ −2 SD; birth length was ≤ −2 SD in 3 and head circumference at birth was −3.5 and −2.9 SD in two patients [11]. Final height was ≤ −2 SD in 6 patients, with values of −3.6 SD at 20 years and −4.3 SD at 19 years in the extreme cases; in the opposite direction, a 17-year-old girl with the t(X;10) translocation had obesity (83.3 kg, more than +4 SD) [11]. Short stature affected all three members of the family of Latchman and colleagues [12]. Microcephaly was recorded in 5/19 in the founding series [11] and in 9/13 (69.2%) in a compilation with different denominators [26], which illustrates how frequencies vary with the criterion and the series. A girl with the p.(Arg14Ter) variant had precocious and rapidly progressive puberty [33].

3.6 Face and eyes

Craniofacial dysmorphism is present in almost all patients (16/18), but it is non-specific and does not constitute a recognisable gestalt [11, 26]. The annotated features include downslanting or upslanting palpebral fissures, epicanthus, hypertelorism, ptosis, a broad or prominent nasal root, a low-hanging columella, a long and smooth philtrum, an accentuated Cupid's bow, a narrow or wide mouth, micrognathia, midface retrusion, low-set ears and a prominent helix [1]. Ocular anomalies are among the most frequent manifestations (11/17 in the founding series) [11]. A review of 27 cases up to May 2024 found ptosis in 35%, myopia in 20%, amblyopia in 16%, hypermetropia in 12% and strabismus in 12%, and described a Canadian boy with bilateral congenital ptosis and blepharophimosis, floppy eyelids, telecanthus and myopia [28]. The 52-year-old patient had high myopia [3]. Glaucoma, retinal coloboma and Duane syndrome have been described in isolated cases [11].

3.7 Skeleton and joints

The distal anomalies are usually mild: short or tapering fingers, brachydactyly, cutaneous syndactyly of the 2nd and 3rd toes (6 patients in the founding series), cone-shaped epiphyses, a broad or valgus hallux and clubfoot [1, 11, 28]. OMIM adds 5th-finger clinodactyly, camptodactyly and a long hallux, records clubfoot in two patients and hand or foot anomalies in at least 12, and notes that cone-shaped phalangeal epiphyses were observed in the few patients who were radiographed, which precludes estimating their frequency [1]. Joint hypermobility affects nearly half of the patients (8/18), and scoliosis was recorded in 3/17 [11]. Subsequent reports added bilateral developmental dysplasia of the hip [22], hand contractures and congenital vertical talus [23] and lower limb deformities with chronic pain in adulthood [3]. The functional work of Li and colleagues proposed that altered proliferation and migration of myoblasts contribute to the joint contractures, a hypothesis based on a single patient [23].

3.8 Heart, kidney, genital tract and digestive tract

Congenital heart disease was recorded in 4/17 patients of the founding series (ventricular septal defect, abnormal mitral valve, patent ductus arteriosus) [11], in a Chinese infant with genitourinary anomalies [27] and in the 52-year-old patient [3]. An abstract presented at a congress in 2025 described identification of the disease in newborns with atrioventricular septal defect, as a neonatal presentation; the content of the abstract was not accessible [35]. Renal and urinary anomalies (8/16) include vesicoureteral reflux, pyelectasis and atrophic kidney; cryptorchidism is frequent in boys (2 in the founding series, bilateral in the 2026 case) and hypospadias is annotated in OMIM [1, 11, 23]. In the OMIM synopsis, each cardiac or urogenital malformation occurred in one or two patients, without forming a consistent pattern [1]. Chronic constipation was described in the 2022 Chinese case [22], and a 2.5-year-old Swedish boy with a de novo frameshift variant had Hirschsprung disease, with no other chromosomal abnormalities or known risk variants for this disease; the authors proposed Hirschsprung disease as an additional manifestation, based on the expression of ZMIZ1 in enteric nervous system progenitors [20, 29]. This is a single case, whose full text was not read.

3.9 Hearing

Hearing loss was recorded in 4/19 patients of the founding series, in 33% of the subset of 21 individuals in the pooled analysis, in one of the children of the Latchman family (sensorineural) and in the Chinese boy reported in 2026 [4, 11, 12, 23].

3.10 Prenatal presentation

In a fetus assessed in Ningbo (China) in March 2025, ultrasound at 30 weeks showed microcephaly, a short femur and intrauterine growth restriction; exome sequencing of amniotic fluid and of the parents' blood identified the c.2633dup p.(Gly879ArgfsTer22) variant de novo [17]. This is the only prenatal diagnosis published in the sources consulted.

3.11 Adulthood

Data on adults are scarce. The founding series included two patients aged 19 and 20 years, both with short stature [11]; the father in the Latchman family, aged 53, had ID and short stature [12, 23]. The oldest patient described, with the de novo p.(Thr300Met) variant, had a low-average IQ, high myopia, craniofacial dysmorphism, genitourinary anomalies, heart disease, lower limb deformities and chronic pain, led a long and active life and died unexpectedly at the age of 52; autopsy showed cortical neuronal heterotopias [3]. The abstract does not state the cause of death, so this case does not allow life expectancy to be estimated or the death to be attributed to the disease. A case of inheritance from an asymptomatic father shows that some adult carriers have no recognised manifestations [13, 26].

3.12 Frequencies in literature reviews

4. Clinical Recognition and Warning Signs

4.1 When to suspect the disorder

The disease has no pathognomonic sign or visually recognisable facies, and its inclusion by Orphanet in the group of non-specific syndromic ID means that the diagnosis is almost always molecular, obtained from an exome, a genome or a neurodevelopmental disorder panel requested for ID, GDD, ASD or congenital anomalies [14, 20]. An automated facial analysis tool applied to the Thai case did not recognise a specific pattern [26]. Clinical suspicion may nonetheless guide the interpretation of a result or the choice of a panel when the elements described in the series occur together [4, 11, 28].

4.2 Warning and emergency signs

No metabolic crises or acute decompensations have been described. The situations requiring prompt action are those of the associated malformations and comorbidities: congenital heart disease in the newborn, feeding difficulties with poor weight gain, epileptic seizures, severe constipation from the neonatal period (which, in the only case with Hirschsprung disease, required this disease to be excluded) and undetected hearing loss, which worsens language delay [11, 29, 35]. In children with limited communication, pain and acute illness may manifest only as a change in behaviour; chronic pain was a relevant complaint in the adult patient [3].

4.3 Newborn screening

There is no biochemical or genomic newborn screening for this disease. Early detection of its manifestations depends on universal newborn hearing screening programmes and on developmental surveillance in primary care.

5. Diagnosis: Criteria, Tests and Algorithms

5.1 Diagnostic criteria

No clinical diagnostic criteria have been published [20]. The diagnosis is established in an individual with a neurodevelopmental disorder by identification of a heterozygous pathogenic or likely pathogenic variant in ZMIZ1, classified according to the ACMG/AMP standards (American College of Medical Genetics and Genomics / Association for Molecular Pathology), or of a chromosomal rearrangement that disrupts the gene or its regulatory region with demonstration of the effect on expression [11, 19, 36].

5.2 Applicable guidelines and algorithm

There is no specific guideline for NEDDFSA, and investigation follows the general recommendations for ID and congenital anomalies [18, 37, 38]. The 2021 ACMG evidence-based guideline, developed with the GRADE method for paediatric patients with one or more congenital anomalies with onset before 1 year of age or with GDD or ID with onset before 18 years, strongly recommends that ES or GS be considered first- or second-tier tests, as they have a higher diagnostic yield than conventional tests [18]. The 2014 American Academy of Pediatrics (AAP) recommendation placed chromosomal microarray (CMA) as the first-line test, replacing the karyotype and subtelomeric fluorescence in situ hybridisation (FISH), together with fragile X testing, and regarded exome sequencing as an emerging technique [37]; the divergence is mainly a matter of timing, and ES/GS is today the route most likely to identify a ZMIZ1 point variant. The 2022 EuroGentest/Solve-RD European recommendations for genome sequencing in the diagnosis of rare diseases stress that this test simultaneously detects single-nucleotide variants, structural variants and repeat expansions [38], which is relevant to the rearrangements affecting ZMIZ1. The proposed algorithm derives from these recommendations and from the limitations described in 5.3.

5.3 Test limitations

The clinical sensitivity of each method for NEDDFSA has not been measured. The proportion of known mechanisms only allows an estimate of what each technique misses: in the founding series, 2 of the 19 patients had balanced translocations, which are detected neither by exome sequencing nor by microarray [11]. In the French study of balanced rearrangements, low-coverage WGS detected 49 of the 55 rearrangements (89%) and enabled a diagnosis in 22 of the 49 (44.8%), with ZMIZ1 among the disrupted genes and among those affected by a position effect [19]; these figures refer to a cohort of rearrangements and not to NEDDFSA. The translocation that disrupts an enhancer 276 kb upstream of the gene does not alter the coding sequence and was interpreted only through the expression study [11]. Synonymous or intronic variants with an effect on splicing, such as c.2610C>T p.(Ser870=) in an exonic splicing enhancer, are easily filtered out as benign by analysis pipelines and require targeted analysis [11]. Parental mosaicism was not documented in the sources read, so its contribution cannot be quantified. A negative result does not exclude the disease; reanalysis of exomes from consanguineous families five years after the initial report changed the result in a clinically relevant way in 28 of 152 families (18%) and reclassified 10 variants previously considered pathogenic or likely pathogenic [39].

5.4 Variant interpretation

The high number of VUS in ClinVar (414 records) and the observed/expected ratio of 0.717 for missense variants in gnomAD indicate that amino acid substitutions are tolerated across much of the protein [2, 9]. Pathogenic missense variants are concentrated in the alanine-rich region (amino acids 280–305) and, in smaller numbers, in the TPR and SP-RING/MIZ domains; a missense variant outside these regions, without demonstrated de novo occurrence, calls for caution [2, 4, 11]. De novo truncating variants are usually classified with the criteria PVS1 (null variant in a gene in which loss of function causes disease), PS2 (confirmed de novo occurrence) and PM2 (absence from population databases), as in the 2026 fetus (PVS1, PS2_Supporting, PM2_Supporting) [17, 36]; the 2022 case, a missense variant in the SP-RING/MIZ domain, was classified as pathogenic with PS2, PM1, PM2, PP2 and PP3 [22]. Application of PVS1 to truncating variants in the last exon, which may escape NMD, should take account of the loss of the TAD [11]. Variants inherited from a healthy parent should not be automatically reclassified as benign, given the incomplete penetrance described [13], nor automatically considered causal when another explanation exists, as in the Thai case with biallelic OTUD6B variants [26].

5.5 Report content

The report should state the gene, the transcript with its version (NM_020338.4), the variant in HGVS nomenclature at the complementary DNA and protein levels, the genomic coordinates with the reference genome version, the zygosity (heterozygous), the ACMG/AMP class with the criteria applied, the parental origin when studied and the method, including the regions not covered and the ability to detect structural variants [8, 36]. For rearrangements, the report should describe the breakpoints at nucleotide level and the demonstration of the effect on expression [19].

5.6 Prenatal and preimplantation diagnosis

Prenatal diagnosis can be performed in two situations: when the familial variant is known, by targeted testing of chorionic villi or amniotic fluid; and when ultrasound shows findings such as microcephaly, a short femur and intrauterine growth restriction, by trio exome sequencing, as in the Ningbo fetus [17]. Preimplantation genetic testing applies to families with a known variant (see section 8).

6. Differential Diagnosis

6.1 Framework

Because NEDDFSA has no distinctive phenotype, the differential diagnosis is in practice that of all syndromic ID, and the distinction is almost always made by genetic testing and not by clinical examination [14, 26]. The founding series placed the disease among the “transcriptomopathies”, the group of syndromes caused by transcriptional regulators that share ID, growth restriction and facial dysmorphism [11]. The closest differentials, by phenotype or by the history of discovery, are OTUD6B-related disorder, PRR12 neuro-ocular syndrome, SMARCA4 Coffin-Siris syndrome and, by isolated features, Williams-Beuren, Mowat-Wilson and Kabuki syndromes.

6.2 OTUD6B-related disorder

Biallelic variants in OTUD6B, which encodes a deubiquitinating enzyme of the OTU (ovarian tumor domain) subfamily, were described in 2017 in 12 individuals from 6 families with ID, epileptic seizures and dysmorphism; carriers of loss-of-function alleles also had GDD, microcephaly, absent speech, hypotonia, growth restriction of prenatal onset, feeding difficulties, brain malformations, congenital heart disease and musculoskeletal abnormalities [40]. OMIM designates the disease IDDFSDA (#617452), a name that, like NEDDFSA, emphasises the distal limb anomalies [41]. The systematic comparison by Phetthong and colleagues shows that the two diseases share ID, GDD, hypotonia, microcephaly, feeding difficulties and distal anomalies in at least two thirds of patients; epilepsy (95% versus 16.7%), arched eyebrows with long palpebral fissures, a prominent nasal root and a broad thumb, sometimes with postaxial polydactyly, favour OTUD6B, whereas ophthalmological abnormalities are more frequent in NEDDFSA [26]. The OMIM description allows two levels of severity to be separated within the 2017 series: the 9 patients from 5 families with truncating alleles had early-onset seizures (mainly generalised tonic-clonic, but also absences, myoclonic and atonic), absent speech even in adolescence and inability to walk in all but one, with microcephaly reaching −6.5 SD and tube feeding in 3; the 3 Turkish siblings of consanguineous parents, homozygous for a missense variant, had mild to moderate ID, acquired speech and walking and seizures starting between 18 months and 8 years [41]. In blood cells from a patient with a truncating allele, loss of 26S proteasome activity, accumulation of ubiquitinated proteins and cytoplasmic inclusions in lymphocytes were observed, and the Otud6b-null mouse dies at the end of gestation or shortly after birth, with intrauterine growth restriction and frequent ventricular septal defects [41]. None of these laboratory findings has been described in NEDDFSA. The OTUD6B phenotype has been compared with Rubinstein-Taybi and Kabuki syndromes, and the Thai case, with periorbital oedema, sagging cheeks and a long smooth philtrum, was referred with suspected Williams syndrome [26]. The distinction has consequences for counselling, because inheritance is recessive in OTUD6B and dominant in ZMIZ1; in the Thai case, the simultaneous presence of variants in both genes showed that a positive result in ZMIZ1 does not obviate complete analysis of the exome [26].

6.3 PRR12 neuro-ocular syndrome

PRR12 was the partner gene of ZMIZ1 in the 2015 t(10;19) translocation, whose phenotype could not be attributed to either of the two genes [24]. In 2021, Chowdhury and colleagues gathered 24 individuals with truncating variants, mostly de novo, and established PRR12 haploinsufficiency as the cause of a disease with a neurodevelopmental disorder in all, structural ocular malformations in 12/24 (50%), hypotonia in 61%, heart disease in 52%, growth delay in 54% and renal anomalies in 35% [42]. OMIM designates it neuroocular syndrome 1 (NOC1, #619539), distinct from NOC2 (#168885) caused by variants in DAGLA, and adds to the Chowdhury series visual impairment in 17 of 22 patients assessed (77%), strabismus in 36% and microcephaly in 29%, as well as the absence of recognisable facial dysmorphism in the available photographs; one of the 24 patients had a 3.35 Mb deletion at 19q13.33–q13.41 that included PRR12 and 145 other genes, which justifies CMA when the exome is negative [43]. The first descriptions, in 2018, noted a stellate iris pattern in all three patients, with iris coloboma in two, and a 2021 study that reanalysed the exomes of 263 undiagnosed probands with microphthalmia, anophthalmia or coloboma found truncating PRR12 variants in 4 families, all with asymmetric ocular involvement and one proband with isolated ocular anomalies, without systemic involvement [43]. In one of these families, mother and daughter were carriers, which shows that NOC1, although usually de novo, can be transmitted [43]. OMIM includes the 2015 t(10;19) translocation in the NOC1 entry and records the authors' hypothesis that the Coffin-Siris-like picture was mainly due to disruption of ZMIZ1; as the abstract of the article allows for haploinsufficiency of one or both genes, attribution of that patient's phenotype remains unresolved [24, 43]. Anophthalmia, microphthalmia and colobomas are more frequent and more severe than the ocular anomalies of NEDDFSA, which are mainly ptosis and refractive errors, with a retinal coloboma described in a single patient [11, 42]. Some patients have no ocular abnormalities, such as an 11-year-old Chinese boy with ID, short stature, mild scoliosis and ADHD, which makes genetic testing the only reliable discriminator [44].

6.4 Coffin-Siris syndrome and other BAFopathies

The patient with the 2015 translocation had a phenotype overlapping that caused by variants in SMARCA4, whose protein BRG1 is a subunit of the nBAF complex with which ZMIZ1 interacts in the mouse brain [24]. Coffin-Siris syndrome shares ID, feeding difficulties, hypotonia and corpus callosum anomalies, but is distinguished by hypoplasia of the nail and distal phalanx of the 5th digit, hypertrichosis and coarse facies [45]. The form caused by SMARCA4 (CSS4, #614609) has particular features: in the first series, of 6 Japanese patients with missense variants or an in-frame deletion (five confirmed de novo), all had hypoplasia or absence of the 5th fingernails, hirsutism, thick eyebrows and long eyelashes, and half had sparse scalp hair, whereas microcephaly, visual and hearing abnormalities and epileptic seizures were variable; the 2014 review of genotype-phenotype correlations noted that the face is less coarse and behavioural abnormalities less frequent than in the forms due to other BAF complex genes [46]. The less marked face brings CSS4 closer to NEDDFSA, which reinforces reliance on molecular diagnosis.

6.5 Williams-Beuren, Mowat-Wilson and Kabuki syndromes

Williams-Beuren syndrome enters the differential because of the face and the recently described interaction between ZMIZ1 and GTF2I, one of the genes in the deleted region, but it is distinguished by elastin arteriopathy, hypercalcaemia and the cognitive and behavioural profile, and it is detected by CMA [23, 32, 47]. It is a contiguous gene syndrome, due to a hemizygous 1.5 to 1.8 Mb deletion at 7q11.23 with about 28 genes, whose frequency was estimated in a Norwegian study at approximately 1 in 7,500 live births, higher than the historical estimate of 1 in 10,000 [48]. Unlike NEDDFSA, in which ZMIZ1 is a single gene and point variants predominate, the recurrent deletion is visible on CMA and rarely requires sequencing. Mowat-Wilson syndrome is the differential to consider in syndromic ID with Hirschsprung disease, as in the Swedish boy with a ZMIZ1 variant; more severe ID, frequent epilepsy, microcephaly and the characteristic face point towards ZEB2 [29, 49]. In a 2003 series, Hirschsprung disease occurred in 67.6% of patients, seizures in 82%, congenital heart disease in 47% and agenesis of the corpus callosum in 35%, which means that about one third of patients with Mowat-Wilson syndrome do not have Hirschsprung disease and are recognised by the face; in a review of 54 brain MRI scans of patients with a confirmed ZEB2 alteration, 96% were abnormal, mainly owing to anomalies of the corpus callosum (79.6%) and the hippocampus (77.8%) and ventricular dilatation (68.5%) [50]. The estimated prevalence is 1 in 50,000 to 70,000 live births, and most variants lead to loss of one functional copy through premature termination or gene deletion, although rare missense variants are associated with attenuated forms [50]. Kabuki syndrome shares ptosis, heart disease, minor skeletal anomalies and mild to moderate ID, but has characteristic facies and persistent fetal fingertip pads [51]. OMIM separates KABUK1 (#147920, KMT2D, autosomal dominant) from KABUK2 (#300867, KDM6A, X-linked); KMT2D variants, mostly truncating and compatible with haploinsufficiency, were found in 74% of 110 families in a 2011 North American series and in 61.7% of 81 patients in a 2013 Japanese series, in which 6.2% had KDM6A variants [52]. Among the malformations, congenital heart disease affected 58% of 60 patients assessed by echocardiography, with coarctation of the aorta as the most frequent defect, and cleft palate was reported in about 40% of a non-Japanese series; IgA deficiency (15 of 19 patients) and hypoglycaemia also point towards Kabuki syndrome [52]. The estimated frequency in Japan is at least 1 in 32,000 [52].

7. Treatments and Risk Management Strategies

7.1 Principles

There is no treatment directed at the cause. The most recent review of ZMIZ1 summarises the current clinical utility of the diagnosis as essentially diagnostic: management is supportive and symptom-driven, and no ZMIZ1-specific intervention is available [20]. There is no guideline, expert consensus or GeneReviews chapter devoted to the disease [5, 20]. Molecular diagnosis nevertheless has practical consequences: it ends the aetiological investigation, allows the recurrence risk to be calculated, guides the active search for the malformations and comorbidities described and connects families with others who have the same diagnosis [20]. Management is multidisciplinary and follows the general principles for ID and GDD, with attention to the manifestations whose frequency in NEDDFSA is documented.

7.2 Management by manifestation

7.3 Pharmacological treatment

No drug has an approved indication or specific evidence for NEDDFSA, and the sources do not describe doses used in patients with ZMIZ1 variants [20]. The only report with detailed pharmacological treatment is that of the Thai child, whose phenotype is attributed mainly to the biallelic OTUD6B variants: febrile and then afebrile seizures (atonic and generalised tonic-clonic) were treated with phenobarbital, discontinued because of a skin reaction, then with sodium valproate, which controlled the seizures, and at 7 years with topiramate after two new seizures; the article does not give doses, and this experience does not constitute evidence for NEDDFSA [26]. Because epilepsy is infrequent in the disease (3/18) and has no characteristic seizure type, the choice of antiseizure medication follows the general recommendations for seizure type and age [11]. The drugs used for irritability associated with ASD and for ADHD are prescribed according to the general indications and labelled conditions; extrapolation of evidence obtained in children with ASD or ID of other causes is the only basis available and is not equivalent to evidence in NEDDFSA. Doses are not presented here because the choice depends on the comorbidity and not on the genetic disease.

7.4 Regulatory status

Source: absence of targeted agents in clinical development according to the 2026 review and absence of trials on ClinicalTrials.gov [20].

7.5 Clinical trials and preclinical research

The ClinicalTrials.gov search (application programming interface version 2, terms “ZMIZ1” and disease name, 28 October 2026) returned no registered study, and the 2026 review confirms that no agent targeting ZMIZ1 has entered clinical development [20]. There are therefore no human trials to tabulate; the table brings together the experimental studies with potential therapeutic relevance.

These results do not allow benefit in patients to be inferred. Rescue by ephrin-B2 was demonstrated in a model of complete forebrain deletion of the gene, which does not reproduce human heterozygosity, and the outcomes are morphological; rescue by the wild-type allele in an overexpression model is not equivalent to gene replacement therapy [11, 30]. The 2026 review adds a safety argument against strategies that reduce ZMIZ1 activity: the dose–response relationship is not monotonic (in pancreatic beta cells, both loss and gain of ZMIZ1 are harmful), a systemic inhibitor would probably have a narrow therapeutic index, with predictable effects on vascular and lymphatic homeostasis, lymphopoiesis and glucose regulation, and exposure during development would be contraindicated [20]. The same search found no completed, suspended or withdrawn trials.

7.6 Developments in the last 12 months

Between November 2025 and October 2026 the following were published: the functional study of the p.(Ala304Pro) variant, with increased ZMIZ1 expression and interaction with GTF2I (laboratory study of one case) [23], and a commentary on this interaction (opinion) [32]; the forebrain deletion mouse model, with cortical microcephaly, corpus callosum dysgenesis and rescue by ephrin-B2 in culture (preclinical) [30], accompanied by an editorial (opinion) [34]; a review of the functions of ZMIZ1 in development and disease (narrative review) [20]; the first published prenatal diagnosis (case report) [17]; and the print version of the report of the 52-year-old patient (case report, published online in October 2025) [3]. There were no regulatory announcements, opinions or authorisations.

7.7 Agents to avoid, pregnancy, perioperative care and transition

The sources do not identify drugs specifically contraindicated in NEDDFSA; the ZMIZ1 inhibitors under investigation in oncology have no application in this disease and exposure to them during development would be contraindicated because of the predictable effects of reducing ZMIZ1 [20]. There are no data on pregnancy or lactation in women with the disease; transmission from an affected father to two children shows that reproduction is possible in adults with a mild phenotype [12]. There are no specific data on anaesthesia; preoperative assessment should take account of congenital heart disease and spinal anomalies when present [11]. In the transition to adulthood, the 52-year-old case points to chronic pain, lower limb deformities and high myopia as problems that require ongoing follow-up [3].

7.8 Guidelines and divergences

In the absence of a specific guideline, the general recommendations apply. For diagnosis, the 2021 ACMG guideline (ES/GS as first- or second-tier tests) diverges from the 2014 AAP recommendation (CMA as first-line test), and the difference reflects the technological evolution between the two dates [18, 37]; the 2022 European recommendations support GS when structural variants are also to be detected [38]. For ASD, the reference is the 2020 AAP guideline on identification, evaluation and management [53]. For variant interpretation, the 2015 ACMG/AMP standards [36]. The sources consulted do not include specific surveillance recommendations, so the follow-up proposed in section 11 derives from the documented frequencies.

8. Reproductive Options and Family Counselling

8.1 Origin of the variants

In the founding series, the variant was shown to be de novo in all families with parents available (14 of 16 families with point variants, excluding the family of three siblings whose parents were not studied), and both translocations were also de novo [11]. Vertical transmission is documented in one family with an affected father and two affected children [12], and inheritance from an asymptomatic parent has been described in the French network, where, whenever the grandparents were studied, the variant had arisen de novo in the healthy parent [13]. The family of three affected siblings without a variant demonstrated in the parents (who were not studied) is compatible with a mildly affected carrier parent or with germline mosaicism, and the sources do not allow a distinction to be made [11, 25].

8.2 Recurrence risks

8.3 Parental testing and cascade testing

Both parents should be tested whenever a variant is identified, for three reasons: it confirms de novo occurrence, which is a pathogenicity criterion; it identifies a mildly affected parent, as recommended by Latchman and colleagues; and it informs the recurrence risk [12, 36]. When a parent is a carrier, testing extends to siblings and other at-risk relatives, and study of the grandparents may show that the variant arose in the parent [13]. Interpretation of a variant inherited from a parent without manifestations requires weighing incomplete penetrance against the hypothesis that the variant is not causal, particularly for in-frame splice variants and for missense variants outside the alanine-rich region, as the Thai case illustrates [26].

8.4 Prenatal and preimplantation diagnosis

When the familial variant is known, prenatal diagnosis is performed by targeted testing of a chorionic villus or amniotic fluid sample, and preimplantation genetic testing is an option for couples in which one member is a carrier. The severity of the phenotype is not predictable from the genotype or from the severity in the parent, which should be discussed before any decision [4, 12]. For a couple with a child affected by a de novo variant, targeted prenatal testing may be offered because of the residual risk of germline mosaicism. When ultrasound shows microcephaly, a short femur and intrauterine growth restriction, trio fetal exome sequencing can identify the disease, as in the Ningbo case, in which the result informed counselling and reproductive guidance [17].

8.5 Counselling of minors and of adults with ID

Testing asymptomatic minor siblings of a carrier is justified when the result changes surveillance or early developmental intervention, because the disease has treatable manifestations in childhood (ophthalmological, auditory, cardiac, renal); when it only informs a future reproductive decision, it should be deferred until the person can decide. In adults with mild ID, such as the father in the Latchman family, reproductive counselling should be adapted to the level of understanding and include the 50% risk and the variability of the phenotype [12].

9. Special Considerations: Population Diversity and Epidemiology

9.1 Frequency

There is no estimate of prevalence or of birth incidence and no population study; the available sources are counts of published cases, which depend on the inclusion criteria and do not measure frequency in the population [5, 14].

The relative frequency among molecular diagnoses of ID gives a local order of magnitude, not a prevalence: in a Shenzhen (China) cohort of 54 children with ID or a neurodevelopmental disorder studied by array comparative genomic hybridisation (CGH) and exome, the diagnostic yield was 61.11% and there was one ZMIZ1 case [22]; in a cohort of 244 ID trios from Galicia (Spain), the exome identified the cause in 64 patients (25.2%), with de novo variants in 61.5% of diagnoses, and one of the cases had a ZMIZ1 variant [4, 54]. ZMIZ1 variants were also identified in a cohort of 42,607 autism cases [4].

9.2 Geographical distribution and populations

The founding series included patients from France, the United States (the majority), the United Kingdom and Poland [11, 23]. Cases were subsequently published from China, Thailand, Mexico, Sweden and Canada, and cases were identified in cohorts from Spain (Galicia), India, Saudi Arabia and Germany (consanguineous families studied in Leipzig) [23, 39, 54–56]. Lu and colleagues described the origins of the patients known in 2022 as Caucasian, Hispanic, North African and Southeast Asian [22]. Because most variants arise de novo, founder populations are not expected, and none has been described [11]. The sources consulted do not include patients from Portugal or Brazil; this absence reflects the sources read and does not demonstrate that there are no diagnosed cases in those countries. The predominance of European, North American and Asian patients follows access to exome sequencing and to case-matching platforms and does not indicate a real difference in frequency.

9.3 Special populations

Data on adults are limited to four people over 18 years of age with a described phenotype (two patients in the founding series, the father in the Latchman family and the 52-year-old patient) and to asymptomatic carrier parents [3, 11–13, 26]. There are no data on pregnancy in affected women. Prenatal reports are limited to one fetus [17].

9.4 European network, resources and patient organisations

ERN ITHACA (European Network for Rare Malformation Syndromes) is the European Reference Network for rare malformation syndromes and ID; it includes a neurodevelopmental disorders working group, the ILIAD registry and the SysNDD database, and lists its member patient organisations [21]. The founding series was assembled with GeneMatcher, MyGene2, DECIPHER and Matchmaker Exchange, which remain the routes for connecting new cases [11]. Resources for consulting the gene include SFARI Gene, ClinGen, GenCC and ClinVar [2, 7, 15, 16]. The sources consulted do not identify a patient organisation dedicated to NEDDFSA.

10. Illustrative Clinical Case

10.1 Presentation

Lu and colleagues described in 2022 a Chinese girl, the second child of a non-consanguineous couple, born at term in October 2016 by caesarean section for breech presentation, weighing 3000 g and without asphyxia; the mother was 36 years old [22]. At 6 months, asymmetry of the skin folds of the lower limbs led to the diagnosis of bilateral developmental dysplasia of the hip by an orthopaedic surgeon. From the first year she had chronic constipation and, from 2017, she was followed in ophthalmology for bilateral low vision and right-eye strabismus, with bilateral amblyopia [22]. After the first year she still did not speak or walk without support and began a programme at a rehabilitation centre; at 2 years she said simple words and walked [22].

10.2 Findings

Examination showed epicanthus, ptosis, upslanting palpebral fissures, telecanthus and hypertelorism, a broad nasal root, an accentuated Cupid's bow, low-set ears, short fingers and toes, tapering fingers and joint hypermobility; she walked independently but could not maintain single-leg stance for more than 3 seconds and fell easily [22]. Visual evoked potentials showed slightly prolonged P100 and N145 latencies bilaterally; hearing, the heart and the urinary tract were normal; hip radiography at 4 years and 7 months showed acetabular angles of about 22° on the left and 25° on the right; two brain MRI scans (2019 and 2021) and three electroencephalograms (2019, 2020 and 2021) were normal [22]. At the assessment at 5 years and 1 month with the WPPSI-IV, the full-scale IQ was 75 (95% confidence interval 70–82; 5th percentile), with index scores of 77 for verbal comprehension, 83 for visuospatial, 79 for perceptual reasoning, 76 for working memory and 71 for processing speed; adaptive behaviour, assessed with the ABAS-II (Adaptive Behavior Assessment System, 2nd edition), had a general index of 77 (social 71, conceptual 84, practical 80), at the borderline level; there was a weak response to her name and hyperactivity [22].

10.3 Molecular diagnosis

Array CGH was normal (46,XX). Trio exome sequencing identified the heterozygous variant NM_020338.4:c.2330G>A p.(Gly777Glu), in exon 20, in the SP-RING/MIZ domain, absent in the parents and the sister, classified as pathogenic with the criteria PS2, PM1, PM2, PP2 and PP3 [22]. Figure 2 of the article shows the pedigree, the hands, the foot and the hip radiographs (https://pmc.ncbi.nlm.nih.gov/articles/PMC9008544/figure/F2/), and figure 3 the molecular analysis (https://pmc.ncbi.nlm.nih.gov/articles/PMC9008544/figure/F3/).

10.4 Discussion

The case brings together the most constant elements of the disease (motor and language delay, non-specific facial dysmorphism, ophthalmological anomalies, short fingers and hypermobility) with a cognitive outcome at the threshold between borderline functioning and mild ID, and shows that MRI and the electroencephalogram may be normal [11, 22]. The diagnosis was obtained by trio exome after a normal array, the sequence recommended before the 2021 ACMG guideline, which now accepts ES as the first test [18, 22]. Hip dysplasia and constipation were added to the spectrum by this report. The rehabilitation started after the first year corresponds to general practice and not to a disease-specific protocol.

11. Prognosis and Long-Term Follow-up

11.1 Prognosis

The cognitive prognosis is highly variable, from severe ID without speech to mild ID, and some patients have a borderline or low-average IQ [1, 3, 22]. There are no longitudinal data that allow the course to be predicted from the genotype; the associations between domain and phenotype described in the pooled analysis are exploratory [4]. The sources consulted do not describe regression or a progressive course. As regards survival, the oldest adult described lived to the age of 52 with an active life, and the affected father in the Latchman family was 53 years old at the time of the report [3, 12, 23]; the unexpected death of the 52-year-old patient has no known cause in the published abstract and does not allow a risk of sudden death to be attributed to the disease [3]. Short stature persists into adulthood in some patients (final height ≤ −2 SD in 6 of the founding series) [11].

11.2 Surveillance

There is no published surveillance protocol for NEDDFSA [20]. The table organises the assessments according to the manifestations whose frequency is documented; the intervals do not derive from a specific guideline and correspond to an assessment at the time of diagnosis, repeated according to the findings and at developmental follow-up visits.

12. Summary and Final Considerations

ZMIZ1-related neurodevelopmental disorder (NEDDFSA, OMIM #618659) is an autosomal dominant disease, almost always de novo, caused by heterozygous variants in ZMIZ1 or by rearrangements that disrupt the gene or an upstream enhancer. ZMIZ1 is a PIAS-family transcriptional coactivator that acts through SUMOylation on the androgen receptor, p53, NOTCH1 and the TGF-β/SMAD pathway, and is required for the migration and polarisation of cortical neurons. The disease mechanism is unresolved: the gene is intolerant to loss of function, but functional studies of missense variants in the alanine-rich region show both reduced and increased protein expression. The clinical picture combines ID or developmental delay, language delay, behavioural abnormalities with ASD and ADHD, hypotonia, feeding difficulties, growth delay, non-specific facial dysmorphism, frequent ocular anomalies, joint hypermobility and mild anomalies of the hands and feet, with heart disease, renal and urinary anomalies, hearing loss and epilepsy in a minority. Expressivity is variable within families and penetrance may be incomplete. The diagnosis is molecular, by exome or genome, with karyotype and genome sequencing reserved for balanced rearrangements that the exome does not detect; missense variants outside the alanine-rich region and variants inherited from healthy parents require careful interpretation. The differential diagnosis includes OTUD6B-related disorder, which is recessive and has almost constant epilepsy, and PRR12 neuro-ocular syndrome, with structural ocular malformations. There is no specific treatment and there are no clinical trials; management is supportive and directed at the manifestations, and therapeutic research is limited to studies in mice and cells. The counts of published cases, between 19 and 36 depending on the criterion, are not equivalent to a prevalence, and data on adults are limited to a few patients, one of whom lived to the age of 52 with an active life.

13. References

1. OMIM® Online Mendelian Inheritance in Man. #618659 NEURODEVELOPMENTAL DISORDER WITH DYSMORPHIC FACIES AND DISTAL SKELETAL ANOMALIES; NEDDFSA. Johns Hopkins University; last edited 17/03/2020. https://omim.org/entry/618659. Accessed 28 October 2026.
2. National Center for Biotechnology Information. ClinVar: ZMIZ1[gene] [Internet]. Bethesda (MD): NLM; [cited 2026-10-28]. Available from: https://www.ncbi.nlm.nih.gov/clinvar/?term=ZMIZ1%5Bgene%5D
3. Rogan S, Gador A, Carroll E, Friedman JM. ZMIZ1-Associated Neurodevelopmental Disorder in a 52-Year-Old Woman. Am J Med Genet A. 2026;200(2):545-556. doi:10.1002/ajmg.a.64243. PMID: 41039966.
4. Cormier KE, Kantor BL, K C R, Blanco-Fernandez X, Galazo MJ. Association of genetic variants, protein domains, and phenotypes in the ZMIZ1 syndromic neurodevelopmental disorder. Front Neurosci. 2025;19:1605762. doi:10.3389/fnins.2025.1605762. PMID: 40529245; PMC12170594.
5. GeneGuidance. ZMIZ1 — gene record and clinical curation [database]. Version of 2026-09-26. Accessed 2026-10-28.
6. OMIM® Online Mendelian Inheritance in Man. *607159 ZINC FINGER MIZ-DOMAIN CONTAINING 1; ZMIZ1. Johns Hopkins University; last edited 17/03/2020. https://omim.org/entry/607159. Accessed 28 October 2026.
7. SFARI Gene. Human Gene Module: ZMIZ1 [Internet]. Simons Foundation; [cited 2026-10-28]. Available from: https://gene.sfari.org/database/human-gene/ZMIZ1
8. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. PMID: 28967166; PMC5765404.
9. Genome Aggregation Database (gnomAD). ZMIZ1 (ENSG00000108175) — gene constraint [Internet]. Broad Institute; [cited 2026-10-28]. Available from: https://gnomad.broadinstitute.org/gene/ENSG00000108175?dataset=gnomad_r4
10. Ensembl. Transcript ENST00000334512 (ZMIZ1-201), GRCh38, release 116 [Internet]. Hinxton: EMBL-EBI. https://rest.ensembl.org/lookup/id/ENST00000334512. Accessed 28 October 2026.
11. Carapito R, Ivanova EL, Morlon A, Meng L, Molitor A, Erdmann E, et al. ZMIZ1 Variants Cause a Syndromic Neurodevelopmental Disorder. Am J Hum Genet. 2019;104(2):319-330. doi:10.1016/j.ajhg.2018.12.007. PMID: 30639322; PMC6369415.
12. Latchman K, Calder M, Morel D, Rhodes L, Juusola J, Tekin M. Autosomal dominant inheritance in a recently described ZMIZ1-related neurodevelopmental disorder: Case report of siblings and an affected parent. Am J Med Genet A. 2020;182(3):548-552. doi:10.1002/ajmg.a.61446. PMID: 31833199.
13. de Masfrand S, Cogné B, Nizon M, Deb W, Goldenberg A, Lecoquierre F, et al. Penetrance, variable expressivity and monogenic neurodevelopmental disorders. Eur J Med Genet. 2024;69:104932. doi:10.1016/j.ejmg.2024.104932. PMID: 38453051.
14. Orphanet. Non-specific syndromic intellectual disability (ORPHA:528084) [Internet]. Paris: INSERM; 2026 [updated 2026-07-02; cited 2026-10-28]. Available from: https://www.orpha.net/en/disease/detail/528084
15. ClinGen. ZMIZ1 — Gene-Disease Validity: complex neurodevelopmental disorder (MONDO:0100038), AD, Definitive (Intellectual Disability and Autism GCEP, 2022-07-05) [Internet]. [cited 2026-10-28]. Available from: https://search.clinicalgenome.org/kb/genes/HGNC:16493
16. The Gene Curation Coalition (GenCC). ZMIZ1 (HGNC:16493) [Internet]. [cited 2026-10-28]. Available from: https://search.thegencc.org/genes/HGNC:16493
17. Zou J, Li H, Yan L. [A fetus with Neurodevelopmental disorders with deformed facial features and distal skeletal abnormalities due to a rare variant of ZMIZ1 gene and literature review]. Zhonghua Yi Xue Yi Chuan Xue Za Zhi. 2026;43(4):295-300. doi:10.3760/cma.j.cn511374-20251129-00693. PMID: 41918386.
18. 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.
19. Schluth-Bolard C, Diguet F, Chatron N, Rollat-Farnier PA, Bardel C, Afenjar A, et al. Whole genome paired-end sequencing elucidates functional and phenotypic consequences of balanced chromosomal rearrangement in patients with developmental disorders. J Med Genet. 2019;56(8):526-535. doi:10.1136/jmedgenet-2018-105778. PMID: 30923172.
20. K C R, Shenoy A, Siddika MA, Patel NR, Meadows SM, Galazo MJ. Decoding Zmiz1: a key transcription coactivator in development and disease. Front Cell Dev Biol. 2026;14:1933470. doi:10.3389/fcell.2026.1933470. PMID: 42923466; PMC13687641.
21. ERN ITHACA. European Network for Rare Malformation Syndromes [Internet]. [cited 2026-10-28]. Available from: https://ern-ithaca.eu/
22. Lu G, Ma L, Xu P, Xian B, Wu L, Ding J, et al. A de Novo ZMIZ1 Pathogenic Variant for Neurodevelopmental Disorder With Dysmorphic Facies and Distal Skeletal Anomalies. Front Genet. 2022;13:840577. doi:10.3389/fgene.2022.840577. PMID: 35432459; PMC9008544.
23. Li S, Zhou D, Han Y, Feng C, Lv X, Fu G, et al. Genetic and functional analysis of ZMIZ1 in neurodevelopmental disorder with dysmorphic facies and distal skeletal anomalies (NEDDFSA): insights from muscle cells and signaling pathways. Pediatr Res. 2026;100(2):872-884. doi:10.1038/s41390-025-04612-x. PMID: 41354990; PMC13638472.
24. Córdova-Fletes C, Domínguez MG, Delint-Ramirez I, Martínez-Rodríguez HG, Rivas-Estilla AM, Barros-Núñez P, et al. A de novo t(10;19)(q22.3;q13.33) leads to ZMIZ1/PRR12 reciprocal fusion transcripts in a girl with intellectual disability and neuropsychiatric alterations. Neurogenetics. 2015;16(4):287-298. doi:10.1007/s10048-015-0452-2. PMID: 26163108.
25. Carapito R, Ivanova EL, Morlon A, Meng L, Molitor A, Erdmann E, et al. ZMIZ1 Variants Cause a Syndromic Neurodevelopmental Disorder [erratum]. Am J Hum Genet. 2020;106(1):137. doi:10.1016/j.ajhg.2019.11.014. PMID: 31879022; PMC7042478.
26. Phetthong T, Khongkrapan A, Jinawath N, Seo GH, Wattanasirichaigoon D. Compound Heterozygote of Point Mutation and Chromosomal Microdeletion Involving OTUD6B Coinciding with ZMIZ1 Variant in Syndromic Intellectual Disability. Genes (Basel). 2021;12(10):1583. doi:10.3390/genes12101583. PMID: 34680978; PMC8535745.
27. He L, Wang Y, Pan J, Guo L, Zhou H, Zhang L. Clinical report and genetic analysis of a novel variant in ZMIZ1 causing neurodevelopmental disorder with dysmorphic factors and distal skeletal anomalies in a Chinese family. Genes Genomics. 2024;46(4):489-498. doi:10.1007/s13258-023-01480-9. PMID: 38117436.
28. Javidi E, Javidi S, Antaki F, Campeau PM, Ospina LH. A novel ZMIZ1 variant associated with NEDDFSA and new ocular features: case report and review of literature. Ophthalmic Genet. 2025;46(1):92-100. doi:10.1080/13816810.2024.2438652. PMID: 39658964.
29. Valind A, Stenström P, Samuelsson S, Jonson T, Eklund EA. ZMIZ1-associated neurodevelopmental disorder and Hirschsprung disease. J Pediatr Surg Case Rep. 2021;71:101889. doi:10.1016/j.epsc.2021.101889.
30. K C R, Patel NR, Thurmon A, Kantor BL, Lorino MG, Tiemroth AS, et al. Loss of Zmiz1 in Mice Leads to Impaired Cortical Development and Autistic-Like Behaviors. Biol Psychiatry. 2026;100(4):447-461. doi:10.1016/j.biopsych.2026.01.014. PMID: 41633496; PMC13474620.
31. K C R, Tiemroth AS, Thurmon AN, Meadows SM, Galazo MJ. Zmiz1 is a novel regulator of brain development associated with autism and intellectual disability. Front Psychiatry. 2024;15:1375492. doi:10.3389/fpsyt.2024.1375492. PMID: 38686122; PMC11057416.
32. De Rosa A, Kalantari S. The ZMIZ1-GTF2I interaction as a potential contributor to the neurodevelopmental phenotype in neurodevelopmental disorder with dysmorphic facies and distal skeletal anomalies (NEDDFSA). Pediatr Res. 2026;100(2):561-562. doi:10.1038/s41390-026-04926-4. PMID: 41981097.
33. Deng S, Xia D, Hou L, Li X, Peng X, Xiao X, et al. A novel ZMIZ1 pathogenic variant in a Chinese girl with neurodevelopmental disorder and review of the literature [preprint]. Research Square. 2022 Oct 17. doi:10.21203/rs.3.rs-2145686/v1.
34. Smith KM. Transcription Factor ZMIZ1 Is a Neurodevelopmental Disorder Risk Gene That Regulates Multiple Steps in Mammalian Brain Development. Biol Psychiatry. 2026;100(4):353-354. doi:10.1016/j.biopsych.2026.06.003. PMID: 42486541.
35. Kruidenier L, Keefe A, Wenger T. P476: Identification of ZMIZ1-related neurodevelopmental disorder in neonates with atrioventricular septal defects: expansion of phenotypes and implications for counseling [conference abstract]. Genet Med Open. 2025;3:103343. doi:10.1016/j.gimo.2025.103343.
36. 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):405-424. doi:10.1038/gim.2015.30. PMID: 25741868; PMC4544753.
37. Moeschler JB, Shevell M, Committee on Genetics. Comprehensive evaluation of the child with intellectual disability or global developmental delays. Pediatrics. 2014;134(3):e903-18. doi:10.1542/peds.2014-1839. PMID: 25157020; PMC9923626.
38. Souche E, Beltran S, Brosens E, Belmont JW, Fossum M, Riess O, et al. Recommendations for whole genome sequencing in diagnostics for rare diseases. Eur J Hum Genet. 2022;30(9):1017-1021. doi:10.1038/s41431-022-01113-x. PMID: 35577938; PMC9437083.
39. Bartolomaeus T, Hentschel J, Jamra RA, Popp B. Re-evaluation and re-analysis of 152 research exomes five years after the initial report reveals clinically relevant changes in 18%. Eur J Hum Genet. 2023;31(10):1154-1164. doi:10.1038/s41431-023-01425-6. PMID: 37460657; PMC10545662.
40. Santiago-Sim T, Burrage LC, Ebstein F, Tokita MJ, Miller M, Bi W, et al. Biallelic Variants in OTUD6B Cause an Intellectual Disability Syndrome Associated with Seizures and Dysmorphic Features. Am J Hum Genet. 2017;100(4):676-688. doi:10.1016/j.ajhg.2017.03.001. PMID: 28343629; PMC5384096.
41. OMIM® Online Mendelian Inheritance in Man. #617452 INTELLECTUAL DEVELOPMENTAL DISORDER WITH DYSMORPHIC FACIES, SEIZURES, AND DISTAL LIMB ANOMALIES; IDDFSDA. Johns Hopkins University; last edited 03/05/2017. https://omim.org/entry/617452. Accessed 28 October 2026.
42. Chowdhury F, Wang L, Al-Raqad M, Amor DJ, Baxová A, Bendová Š, et al. Haploinsufficiency of PRR12 causes a spectrum of neurodevelopmental, eye, and multisystem abnormalities. Genet Med. 2021;23(7):1234-1245. doi:10.1038/s41436-021-01129-6. PMID: 33824499.
43. OMIM® Online Mendelian Inheritance in Man. #619539 NEUROOCULAR SYNDROME 1; NOC1. Johns Hopkins University; last edited 30/04/2024. https://omim.org/entry/619539. Accessed 28 October 2026.
44. Liu Z, Ding S, Xu G, Fang C. Case Report: Identification of a novel PRR12 variant in a Chinese boy with developmental delay and short stature. Front Pediatr. 2024;12:1367131. doi:10.3389/fped.2024.1367131. PMID: 38798311; PMC11119739.
45. Schrier Vergano S, Santen G, Wieczorek D, Matsumoto N. Coffin-Siris Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 23556151. Available from: https://www.ncbi.nlm.nih.gov/books/NBK131811/
46. OMIM® Online Mendelian Inheritance in Man. #614609 COFFIN-SIRIS SYNDROME 4; CSS4. Johns Hopkins University; last edited 06/05/2016. https://omim.org/entry/614609. Accessed 28 October 2026.
47. Morris CA. Williams syndrome. 1999 Apr 9 [updated 2023 Apr 13]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301427. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1249/
48. OMIM® Online Mendelian Inheritance in Man. #194050 WILLIAMS-BEUREN SYNDROME; WBS. Johns Hopkins University; last edited 04/09/2024. https://omim.org/entry/194050. Accessed 28 October 2026.
49. Adam MP, Conta J, Bean LJH. Classic Mowat-Wilson syndrome. 2007 Mar 28 [updated 2025 Apr 10; revised 2026 Jun 30]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301585. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1412/
50. OMIM® Online Mendelian Inheritance in Man. #235730 MOWAT-WILSON SYNDROME; MOWS. Johns Hopkins University; last edited 14/11/2023. https://omim.org/entry/235730. Accessed 28 October 2026.
51. Adam MP, Hannibal M. Kabuki syndrome. 2011 Sep 1 [updated 2019 Feb 28; revised 2026 Apr 23]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 21882399. Available from: https://www.ncbi.nlm.nih.gov/books/NBK62111/
52. OMIM® Online Mendelian Inheritance in Man. #147920 KABUKI SYNDROME 1; KABUK1. Johns Hopkins University; last edited 16/07/2025. https://omim.org/entry/147920. Accessed 28 October 2026.
53. Hyman SL, Levy SE, Myers SM, COUNCIL ON CHILDREN WITH DISABILITIES, SECTION ON DEVELOPMENTAL AND BEHAVIORAL PEDIATRICS. Identification, Evaluation, and Management of Children With Autism Spectrum Disorder. Pediatrics. 2020;145(1):e20193447. doi:10.1542/peds.2019-3447. PMID: 31843864.
54. Brea-Fernández AJ, Álvarez-Barona M, Amigo J, Tubío-Fungueiriño M, Caamaño P, Fernández-Prieto M, et al. Trio-based exome sequencing reveals a high rate of the de novo variants in intellectual disability. Eur J Hum Genet. 2022;30(8):938-945. doi:10.1038/s41431-022-01087-w. PMID: 35322241; PMC9349217.
55. Sheth F, Shah J, Jain D, Shah S, Patel H, Patel K, et al. Comparative yield of molecular diagnostic algorithms for autism spectrum disorder diagnosis in India: evidence supporting whole exome sequencing as first tier test. BMC Neurol. 2023;23(1):292. doi:10.1186/s12883-023-03341-0. PMID: 37543562; PMC10403833.
56. Alqahtani AS, Alotibi RS, Aloraini T, Almsned F, Alassali Y, Alfares A, et al. Prospect of genetic disorders in Saudi Arabia. Front Genet. 2023;14:1243518. doi:10.3389/fgene.2023.1243518. PMID: 37799141; PMC10548463.