DYRK1A Syndrome (Including the 21q22.13 Microdeletion)

Literature and regulatory information updated: 31 October 2026.

Autosomal dominant intellectual developmental disorder 7 (MRD7), caused by haploinsufficiency of the DYRK1A gene through intragenic loss-of-function variants, translocations that disrupt the gene, or 21q22.13 microdeletions that include it.

0. Initial Reference Table

1. Introduction

1.1 Definition and scope

DYRK1A syndrome is an autosomal dominant neurodevelopmental disorder caused by loss of one functional copy of the DYRK1A gene, located at 21q22.13. Orphanet defines it as a rare syndromic ID characterised by microcephaly, global developmental delay, mild to severe ID, speech impairment, feeding problems, behavioural changes (frequently ASD) and facial dysmorphism with prominent ears, deep-set eyes, a short nose with a broad tip and retrognathia with a broad chin, together with, more variably, seizures, short stature and ocular, cardiac, urogenital and musculoskeletal anomalies [12]. The same Orphanet entry (ORPHA:464306) groups two subtypes: the point mutation form (ORPHA:464311) and the 21q22.13q22.2 microdeletion form (ORPHA:268261), described as global delay with severely delayed or absent speech, moderate to severe ID, behavioural changes, stereotypies, febrile seizures and epilepsy, abnormal gait, visual defects, characteristic face, IUGR and feeding difficulties [3, 4]. This article covers both forms together, because the phenotype of microdeletions that include DYRK1A cannot be clearly distinguished from that of intragenic variants, with the caveat that larger deletions involving other bands may have more severe pictures [2, 8].

The OMIM name is "intellectual developmental disorder, autosomal dominant 7" (MRD7, number 614104), which replaced the old designation "mental retardation, autosomal dominant 7"; in the clinical literature and patient associations, "DYRK1A syndrome" or "DYRK1A haploinsufficiency syndrome" prevails [1, 21]. The value 268261, sometimes given as an OMIM number, is the Orphanet code for the microdeletion subtype; the correct numbers are 614104 (phenotype) and 600855 (gene) [1, 4, 5].

1.2 The gene before the syndrome

DYRK1A was cloned in the mid-1990s in the 21q22 region homologous to the segment of mouse chromosome 16, as the human homologue of the Drosophila minibrain (mnb) gene, whose mutants have an adult brain 40–50% smaller [5, 22]. Its location in the so-called Down syndrome critical region made it, from the outset, a candidate gene for the learning deficits of trisomy 21: in transgenic mice carrying extra copies of 21q22.2 fragments, the learning and memory deficits were mapped to a 180 kb interval containing the minibrain homologue [5]. This history explains why DYRK1A syndrome and Down syndrome are described as the opposite ends of a dosage axis of the same gene: one copy in the former, three copies in the latter [6, 23, 24].

1.3 Milestones in the clinical and molecular description

In 2008, Møller and colleagues described two unrelated patients with prenatal-onset microcephaly, IUGR, postnatal feeding difficulties, developmental delay and febrile seizures, each with a de novo balanced translocation disrupting DYRK1A, t(9;21)(p12;q22) and t(2;21)(q22;q22); the authors noted the similarity to partial monosomy 21 and to Dyrk1a+/− mice, which have a brain about 30% smaller, growth retardation and behavioural and motor changes [1, 25]. In 2011, van Bon and colleagues identified a 52 kb de novo intragenic deletion, removing the last three exons, in a woman with primary microcephaly (−3 standard deviations, SD), severe ID without speech and anxious autistic behaviour, in a screen for DYRK1A copy-number variants (CNVs) in 3 009 people with ID [5]. In 2012 the first point variants appeared: O'Roak and colleagues found three de novo variants (two frameshift and one splicing) among 2 446 probands with ASD, all with a head circumference (HC) smaller than that of the other screened individuals and smaller than that of their parents, and Courcet and colleagues described a de novo 2-nucleotide deletion in a 14-year-old girl with severe ID, microcephaly of −6 SD and epilepsy, the only carrier among 150 patients with a similar phenotype [1, 5, 13].

Between 2015 and 2017, larger series delineated the syndrome. Ji and colleagues brought together 14 patients (five microdeletions, three insertions/deletions and six single-nucleotide variants) and showed a consistent phenotype across all types of alteration, with congenital microcephaly of −2 to −5 SD, ID, severely impaired speech, short stature and a distinctive face [8]. Bronicki and colleagues gathered ten patients and suggested looking for DYRK1A in ID with microcephaly, feeding difficulties and absent or delayed speech [26]. Van Bon and colleagues resequenced the gene in 7 162 patients with ASD or developmental delay and demonstrated an excess of truncating and de novo variants (p = 2.53×10⁻¹⁰), whereas 6 503 controls had no truncating variants [22]. Earl and colleagues compared 61 patients with almost 2 000 children with idiopathic ASD and described the clinical profile that serves as the reference today [16]. GeneReviews brought together, in its 2021 update, 68 published individuals [2].

1.4 From chromosomal microdeletion to monogenic disease

DYRK1A syndrome illustrates the blurring of the boundary between microdeletion syndromes and monogenic disorders. The Oxford Desk Reference stresses that many microdeletions owe their phenotype to a dosage change of a single gene within the interval, and that a "chromosomal phenotype" should now lead to broad genetic investigation (microarray, panel, exome or genome) [27]. Smith's Recognizable Patterns of Human Malformation adds that submicroscopic deletions or duplications removing a single gene can be considered monogenic disorders and that FISH, multiplex ligation-dependent probe amplification (MLPA) and arrays identify them [28]. In the eight reference textbooks consulted, DYRK1A syndrome has no entry or specific mention of its own; these books provide the framework for the group (ID, microcephaly, ASD and microdeletions) used throughout this article [11, 29, 30].

2. Molecular Genetics and Pathophysiology

2.1 Gene structure and isoforms

DYRK1A has 17 exons, with alternative use, spanning about 150 kb, and two promoters in the 5′ untranslated region: one ubiquitous (MNBHa transcript) and the other active mainly in the heart and skeletal muscle (MNBHb) [5]. Alternative splicing generates at least four isoforms differing at the C-terminus; the most abundant isoform in the brain (isoform 1) has 763 amino acids and about 85.6 kDa and contains a repeat of 13 histidines and a serine/threonine-rich domain, whereas the isoform initially cloned has 754 amino acids, 9 fewer near the N-terminus, through use of an alternative splicing acceptor [5]. The protein has a nuclear localisation signal, a kinase domain and a possible leucine zipper, and is more than 99% identical between human, mouse and rat [5].

This duality has practical consequences for nomenclature. The MANE Select transcript (Matched Annotation from NCBI and EMBL-EBI, the reference transcript agreed between the North American and European databases), which is also the RefSeq Select, is NM_001347721.2, encoding the 754-amino-acid isoform (NP_001334650.1), but most of the clinical literature, including GeneReviews, the French series and case reports, uses NM_001396.3, .4 or .5, encoding the 763-amino-acid isoform (NP_001387.2) [7, 10, 22]. Positions downstream of the 9-amino-acid segment differ by 27 nucleotides and 9 codons between the two transcripts, and the conversions presented in this article, with coordinates on the GRCh38 reference genome assembly, were all confirmed in VariantValidator [7]. The same segment has clinical significance: truncating variants in the region encoding amino acids 70–78 of the NM_001396.4 isoform, present only in some isoforms, have been described in people without ID and are considered non-pathogenic [13].

2.2 Protein function

DYRK1A autophosphorylates a tyrosine residue in the activation loop during its synthesis and then phosphorylates substrates on serine and threonine [6]. Among the substrates and partners described are the C-terminal domain of RNA polymerase II, RNF169 (which controls accumulation of TP53BP1 at double-strand breaks and the choice of repair by homologous recombination), the splicing factors SRSF6 and SF3B1, SIRT1, septins 4 and 5, tau protein, the NFATc transcription factors (in synergy with RCAN1), REST/NRSF, MAP1B and caspase-9 in the retina [5, 6]. The protein is highly expressed in the developing central nervous system and regulates neurogenesis, neuronal proliferation and differentiation, synaptic transmission and neurodegenerative processes [2]. It also acts as a circadian clock kinase, through CRY2, and interacts with the SWI/SNF chromatin remodelling complex, which offers a plausible basis for the sleep disturbances and for the overlap with other chromatin gene syndromes [22].

2.3 Mechanism: haploinsufficiency

The mechanism is haploinsufficiency through inactivation of one allele [2]. ClinGen gives DYRK1A the maximum haploinsufficiency score (3) and no evidence of triplosensitivity (0), and the probability of loss-of-function intolerance (pLI) is 1 [6, 31]. Causal alterations include balanced translocations disrupting the gene, intragenic or whole-gene deletions, nonsense, frameshift and splicing variants, and missense variants that abolish kinase activity, most clustered near the adenosine triphosphate (ATP)-binding pocket and the catalytic centre of the kinase domain [2, 10]. Animal models reproduce the disease: Dyrk1a+/− mice have reduced viability, pre- and postnatal growth retardation, a brain about 30% smaller, anxiety and dopaminergic and motor changes; the eyes are 25% smaller, the retina is thinner, there are fewer ganglion cells and the electroretinogram (ERG) is abnormal [17, 22, 25].

More recent experimental work adds circuit mechanisms. In mice, Dyrk1a regulates the tangential migration of GABAergic interneurons in a manner dependent on dosage and kinase activity, through the actomyosin cytoskeleton; heterozygous inactivation restricted to interneurons reduces their cortical density and produces behavioural changes and epileptic activity, a model of excitation–inhibition imbalance [32]. Knock-in mice carrying the human variant p.(Ile48LysfsTer2) have severe microcephaly, dendritic retraction, reduced excitatory transmission and social deficits [33]. In mutant mice, synaptic genes in the prefrontal cortex are downregulated [34]. Murine models also show lower bone mineral density of the femur, mandible and skull, fewer trabeculae and lower mechanical strength, with earlier deficits in males, without equivalent studies in patients [21].

2.4 Variant spectrum and normalisation

Sequence variants account for about 87% of the recorded pathogenic alterations and gene-targeted deletions or duplications for about 13%; this proportion describes the catalogued variants, not the clinical sensitivity of each test [2]. In the review of 108 variants by Méjécase and colleagues, 75% were single-nucleotide variants or small insertions/deletions (40 nonsense and 39 frameshift) and 25% chromosomal rearrangements [17]. There are no founder variants; most are private, with some recurrent codons. The following table presents variants cited in this article, normalised on both transcripts.

The missense variants studied functionally by Courraud and colleagues are numbered on the 763-amino-acid isoform (NP_001387.2). The variants p.Gly168Asp, p.Asp287Val, p.Ile305Arg, p.Ser311Phe and p.Arg467Gln reduced the amount of protein and abolished autophosphorylation of tyrosine 321, p.Ser324Arg had a partial effect, and the variants p.Arg255Gln, p.Tyr462His, p.Gly486Asp and p.Thr588Asn had no effect in that assay [10]. Very distal truncating variants, which escape nonsense-mediated messenger RNA decay (NMD) and preserve the kinase domain, call for caution: p.Ser660* and p.Ser661* had no effect in the autophosphorylation assay, although the frameshift p.Ser660fs has a positive methylation signature [10, 13].

2.5 Genotype–phenotype correlation, penetrance and expressivity

No consistent genotype–phenotype correlations have been identified; the series of Ji and of Earl found no significant differences between intragenic variants and chromosomal rearrangements [2, 8, 16]. Penetrance of de novo loss-of-function variants is considered to be about 100%, haploinsufficiency has not been observed in control populations and expressivity is similar in both sexes [2, 22]. Expressivity is, however, more variable than the first series suggested. In the University of Washington cohort, three participants did not have ID: a 5-year-old boy with a de novo missense variant (full-scale intelligence quotient, IQ, 87), a 4-year-old boy with a de novo frameshift (IQ 68) and a 14-year-old boy with an inherited pathogenic frameshift and IQ 133, with ASD, depression, developmental coordination disorder, dysmorphism, pectus, scoliosis, seizures, constipation and inguinal hernia [9]. In the same programme, a carrier parent had IQ 102 and elevated autistic traits, and Earl and colleagues showed that the variability of parental HC, IQ and ASD symptoms tracks that of the children, which indicates an effect of the familial genetic background on severity [9, 16]. Van Bon and colleagues described a splicing variant inherited from a healthy mother, c.208-1G>A, affecting only three of the four isoforms, in a patient with normal IQ and without microcephaly, and considered it probably inconsequential, an example of a predicted loss-of-function variant that should not be interpreted without caution [22].

2.6 Allelic disorders and somatic alterations

GeneReviews recognises no other germline phenotypes associated with DYRK1A [2]. The variant p.Gly486Asp, with no effect in the autophosphorylation assay, produced a methylation profile opposite to that of the syndrome, suggestive of gain of function, in a person with relative macrocephaly and ASD without ID, which remains an isolated observation [10]. Outside the germline, analysis of 1 386 resected epileptogenic lesions showed enrichment of somatic DYRK1A missense variants in malformations of cortical development, including focal cortical dysplasia type II (dN/dS ratio 27.52; p = 1.60×10⁻⁵), an association distinct from the germline syndrome [40]. At the opposite pole of dosage, trisomy of 21q22 confers much of the Down syndrome phenotype, and DYRK1A overexpression is one of the candidate mechanisms for the learning disability, bone changes and early-onset Alzheimer disease of that syndrome [6, 23, 24].

3. Clinical Spectrum and Manifestations by System

3.1 Data sources and limits

The clinical data come from cumulative literature series and from cohorts recruited by reference centres or by ASD studies, each with its own denominators. The main ones are the review of 68 patients published up to 2021 (GeneReviews), the series of 61 patients of Earl and colleagues (42 with sequence variants and 19 with rearrangements), the cohort of 50 patients of Courraud and colleagues (France, Denmark and Switzerland), the French cohort of 80 patients compared with 32 patients of other origins in the PNDS (French National Diagnosis and Care Protocol), the prospective Australian cohort of 38 patients of Morison and colleagues, the University of Washington cohort of 29 patients of Kurtz-Nelson and colleagues and the ophthalmological review of 145 patients of Méjécase and colleagues [2, 9, 10, 13, 16, 17, 41]. None of these series is population-based; those recruited through ASD tend to overrepresent autism, and those from clinical genetics tend to overrepresent microcephaly and severe ID. Almost all the data describe children and adolescents; adults are few and descriptions of natural history in adulthood are scarce [2, 38]. As there is no disease-modifying treatment, the data describe natural history under usual supportive care.

3.2 Frequency of manifestations

Sources for the table: [2, 13, 16]. The differences between columns reflect definitions (for example, "epilepsy" including or excluding febrile-only seizures), recruitment methods and the thoroughness of assessment; scoliosis and chest changes, for example, vary between about 10% in the literature review and about half in the PNDS, probably because systematic examination in a followed cohort detects mild forms that case reports omit [2, 13].

3.3 Prenatal and neonatal period

Microcephaly, IUGR and oligohydramnios may be detected before birth [2]. HC at birth lies between −1 and −4 SD, and low birth weight (< −2 SD) is frequent [2]. The PNDS notes that microcephaly, brain or urinary tract malformations and IUGR may lead to prenatal genetic testing, but that it is difficult to suggest the syndrome specifically from ultrasound signs [13]. In the neonatal period, sucking and swallowing difficulties and gastro-oesophageal reflux (GOR) predominate, present in more than 80% of patients [2, 13]. In one case, the presentation included suspected Hirschsprung disease with colostomy at 48 hours of life [44].

3.4 Growth and head circumference

Microcephaly is the most constant sign. HC, between −1 and −4 SD at birth, grows slowly and drifts to −2 to −5 SD in most patients [2]. The PNDS describes microcephaly in all patients, inconstant at birth and progressive after birth [13]. In the subcohort of ten patients assessed in Washington, the mean HC z-score was −3.62 (SD 2.17), compared with +1.21 in idiopathic ASD and +1.76 in CHD8-related ASD, and the variant was associated with a reduction of 2.93 SD relative to parental HC [16]. There are exceptions: one patient went from −4.5 SD at 4 years to −1.6 SD at 17 years, another had HC −1.6 SD at 13 years, and a 63-year-old woman had a normal HC (56.8 cm) [5, 22, 38].

After the neonatal period, low weight and a slender build persist; in the GeneReviews review only four individuals had weight above the 50th centile and only two height above the 50th centile [2]. Short stature (< −2 SD) affects about 44%, with prenatal onset or later in childhood [2]. Lanvin and colleagues constructed specific growth charts for height, weight, body mass index (BMI) and HC from 0 to 5 years from 92 patients (49 male and 43 female): 76 taken from the international literature and 16 French patients already included in the cohort of Courraud and colleagues; 67 had truncating variants and 25 non-truncating variants (missense or in-frame deletions) [43]. Mean gestational age was 40 weeks and 3 days. At birth, mean weight was 2750 g in boys (1st centile) and 2480 g in girls (below the 1st centile), mean length 48.2 cm and 45.8 cm and mean HC 32.3 cm and 31.2 cm; 14 of 67 newborns (21%) had weight below −2 SD, 6 (12% of those with weight and length recorded) met the definition of IUGR used by the authors (weight and length < −2 SD) and only 8% had weight above the 50th centile [43]. About two thirds had HC < −2 SD at birth, and after 2 years HC was always below −2 SD; at 5 years, mean height was −1.5 SD in boys and −1.2 SD in girls, mean weight −1.2 SD and −1.4 SD and mean HC −3 SD and −3.2 SD [43]. In the sample, 54.5% had short stature (< −2 SD) and only four individuals height above the 50th centile; growth retardation was homogeneous across variant types, with no genotype–phenotype correlation [43]. The charts end at 5 years owing to the scarcity of later measurements, and the data combine paediatric records, parent-reported measurements and published cases, which limits their precision [43]. The PNDS charts, based on 90 patients (41 girls and 49 boys), end at 9–10 years for weight, height and BMI and at 6 years for HC, owing to lack of data at later ages [13].

3.5 Development, cognition and language

Global delay is constant. Generalised hypertonia may be noticed in the first months, often with associated axial hypotonia; most children walk after 2–3 years, some at the upper limit of normal [2, 13]. In the Washington subcohort, mean age at independent walking was 19.7 months (SD 5.6), compared with 13.2 months in idiopathic ASD, and mean full-scale IQ was 45.3 (SD 18.1), compared with 81.2 [16]. Most function at the level of moderate to severe ID; mild ID and, rarely, normal intellectual functioning have been described [2, 9]. In comparison with Wiedemann-Steiner syndrome, patients with DYRK1A syndrome had lower adaptive behaviour and more affected language and communication [42]. In the Simons Searchlight registry, the 24 participants assessed with standardised scales all had neurological manifestations, of variable severity [45].

Language is the most affected domain. All have speech delay; some learn to speak, others do not speak or use one- or two-word utterances, and expressive language is more affected than receptive language [2]. In the Australian cohort of 38 patients (23 male; median 8 years and 3 months, range 1 year and 7 months to 25 years), only 8 had sufficient oral language to rely on verbal communication alone; in those who spoke, apraxia of speech and dysarthria were severe and greatly reduced intelligibility; 30 combined gestures, signs and graphic augmentative and alternative communication (AAC); expressive, receptive and written language was low; and social motivation stood out as a relative strength in an altered social profile in 25 of 29 assessed [41]. The PNDS stresses that very limited verbal ability does not reflect comprehension and that communicative frustration contributes to behavioural changes [13].

3.6 Behaviour and mental health

ASD, stereotypies, anxiety, hyperactivity and sleep disturbances (difficulty falling asleep and night waking) are frequent [2]. With standardised assessment (Autism Diagnostic Observation Schedule, ADOS-2, and Autism Diagnostic Interview-Revised, ADI-R), Kurtz-Nelson and colleagues confirmed ASD in 23 of 27 patients (85.2%), with difficulties in social reciprocity and non-verbal communication, high sensory-seeking behaviours and fewer idiosyncratic negative sensory responses than in idiopathic ASD [9]. In Earl's series, 89% of patients identified through ASD had five or more of the cardinal symptoms of the syndrome [16]. The psychiatric profile is heterogeneous, with anxiety, depression, attention deficit hyperactivity disorder (ADHD) and oppositional behaviour, unrelated to age at first word, age at walking or current age [46]. The study of executive function draws attention to the risk of attributing all symptoms to ID and of failing to diagnose ADHD or executive deficits in non-verbal people, which requires adapted measures [47].

3.7 Epilepsy and neuroimaging

Febrile seizures are common in infancy, and about half of patients develop epilepsy, with atonic seizures, absences and generalised myoclonic seizures; some children have only febrile seizures [2]. The PNDS describes seizures in more than half, beginning in most with febrile seizures, with polymorphic semiology, mainly generalised (atonic, absences, tonic-clonic or myoclonic), and epilepsy generally controlled with treatment [13]. In van Bon's series, 77% had febrile seizures and 33% later epilepsy [22]. The review by Okazaki and colleagues reports that epilepsy tends to be controlled, with reports of treatment with valproate and levetiracetam [36].

MRI shows signs of global brain underdevelopment or hypomyelination: enlarged ventricles, delayed myelination, cortical atrophy, hypoplasia of the corpus callosum, a small brainstem and a hypoplastic pituitary stalk [2]. The PNDS finds non-specific changes in about 80% and does not recommend systematic MRI [13]. Okazaki and colleagues described hyperintensities on T2 and fluid-attenuated inversion recovery (FLAIR) sequences in the frontal and parietal white matter, sparing the U-fibres, which decreased between 11 and 15 years of age [36].

3.8 Motor function and abnormal movements

Gait is typically broad-based or ataxic [2]. Earl and colleagues described a "waddling" gait, with the trunk leaning forward, arms flexed and held close to the body, and open hands [16]. Limb hypertonia with axial hypotonia is characteristic in childhood [13]. The adult motor spectrum is poorly defined. Santos Sánchez de las Matas and colleagues described a 63-year-old woman with severe ID and autism whose gait and manipulation were normal until the third decade, when progressive twisting postures appeared and evolved into severe multifocal dystonia with fixed contractures of the neck and limbs, with jaw-opening stereotypies and tonic upward gaze deviations since childhood [38].

3.9 Eyes

Ocular anomalies are frequent and potentially treatable. In the review by Méjécase and colleagues, which brought together 112 patients from the literature, 19 from the UK 100 000 Genomes Project and responses to a DSIA survey, 90 of 145 patients (62.1%) had ocular findings, 18.6% did not and information was missing in 19.3%; among those affected, 35.6% had refractive error, 21.1% strabismus, 25.6% enophthalmos and 13% optic nerve hypoplasia, and seven had retinal involvement, including detachment [17]. In the survey, the 14 patients with molecular confirmation all had strabismus, 8 of them exotropia [17]. The PNDS adds sixth nerve palsy, retinal hypopigmentation, optic disc pallor, microphthalmia and, rarely, blepharophimosis, ptosis, bluish sclerae, cataract and coloboma, with visual evoked potentials and ERG rarely abnormal [13]. In a series of five patients from South Korea, all had exotropia, four optic atrophy and two amblyopia [48]. In a 9-year-old Turkish boy with the de novo variant NM_001396.5:c.478G>T p.(Glu160Ter) (NM_001347721.2:c.451G>T p.(Glu151Ter)), strabismus at 9 months led to the diagnosis of hypermetropia and optic atrophy; the same patient had an HC of −5.84 SD, toe walking from 13 months, no progression of speech, stereotypies, self-injury and, on echocardiography, mild subvalvular aortic stenosis, aortic and mitral regurgitation and tortuosity of the aorta [7, 15].

Retinal involvement, present in about 5% of patients, deserves specific attention [44]. Cai and colleagues described a 4-year-old girl with iris coloboma, lattice degeneration and avascularity of the peripheral retina, with vascular pruning and leakage on wide-field fluorescein angiography, treated with laser photocoagulation of the avascular retina because of the risk of detachment [49]. Lin and colleagues described a retinovascular phenotype similar to that of familial exudative vitreoretinopathy (FEVR), with microphthalmia, a retinal fold and total detachment in one eye and macular dragging in the other, and proposed that these changes may serve as a diagnostic clue [44].

3.10 Face, skin and teeth

In infancy and childhood, the face shows prominent ears, deep-set eyes, slightly upslanting palpebral fissures, a short nose with a broad tip and retrognathia with a broad chin; in adults, the nasal bridge is high, the alae nasi underdeveloped and the nose more prominent [2]. Earl and colleagues added a hooded appearance of the eyes, bitemporal narrowing, a prominent forehead with a high hairline, a tubular nose, a small chin and thick, overfolded helices [16]. The PNDS describes prominent, poorly hemmed ears, upper eyelid oedema, a thin upper lip, thin skin and fine hair with an irregular hairline [13]. Ethnicity alters facial appearance and may make the syndrome less recognisable, as shown by an Italian series of four patients of different origins [50]. Atopic or thin and translucent skin is present in about two thirds of the French cohort, with atopic dermatitis highlighted by Courraud and colleagues and eczema already mentioned in the 2011 case [1, 10, 13]. Dental anomalies include widely spaced teeth, extreme dental calculus, delayed primary dentition, neonatal teeth and supernumerary teeth [2].

3.11 Heart, digestive tract and urogenital tract

Congenital heart defects affect 9 of 48 patients in the GeneReviews review and about 15% in the PNDS: atrial septal defect, ventricular septal defect, patent ductus arteriosus, aortic and pulmonary valve anomalies, aortic stenosis, branch pulmonary stenosis and hypoplastic left heart [2, 13]. Van Bon and colleagues found cardiac anomalies in 18% and ophthalmological anomalies in 27% of their series and recommended assessment of both in all patients [22]. In the digestive tract, GOR and constipation, sometimes severe, predominate [2]; rare anomalies have been described, such as splenic torsion with abnormal intestinal fixation in a 17-year-old girl [51].

Congenital anomalies of the kidney and urinary tract (CAKUT) and genital defects justify systematic screening. In the Baylor Genetics clinical exome database (about 8 000 probands), 19 patients had DYRK1A variants; of the 15 with information, 11 (73%) had CAKUT and/or genital defects and 36% renal anomalies, such as unilateral renal agenesis; in Xenopus, knockdown of Dyrk1a disrupted segments of the embryonic nephron, a defect rescued by normal human DYRK1A but not by the variants p.R205* and p.L245R [31]. GeneReviews lists cryptorchidism, hypoplastic or shawl scrotum, micropenis, hypospadias, inguinal hernia, frequent urinary tract infections, vesicoureteric reflux and unilateral renal agenesis [2]. In the cohort of Courraud and colleagues there were genital anomalies but no evident renal anomalies, which shows how these frequencies depend on the extent of investigation [10].

3.12 Skeleton, endocrinology and other features

In addition to scoliosis, kyphosis and pectus excavatum, the foot anomaly is typical: mild cutaneous syndactyly of toes 2–4, hallux valgus and a short fifth toe [2]. Contractures of the ankles, feet and hands, pes cavovarus, tibial osteochondrosis, pes cavus, a long hallux and arachnodactyly have been described [16, 21]. There are no studies of bone mineral density in patients, and the hypothesis of bone fragility rests only on murine models [21]. The PNDS mentions growth hormone (GH) deficiency, hypothyroidism and precocious puberty as possible findings, without frequencies [13].

3.13 21q22.13 microdeletions

Deletions that include DYRK1A produce mild to severe developmental delay, speech impairment, ataxic-type gait, short stature, low weight, seizures and a characteristic face, with no clear difference from intragenic variants [2]. In the series of Ji and colleagues, the five 21q22.12q22.3 microdeletions had the same picture as the sequence variants [8]. Orphanet describes for this subtype severely delayed or absent speech, moderate to severe ID, stereotypies, febrile seizures and epilepsy, visual defects, IUGR and feeding difficulties [4]. Larger deletions involving other bands may have more severe phenotypes, assessable case by case in databases such as DECIPHER [2]. Taşdelen and colleagues described a Turkish female infant, the daughter of first cousins, with IUGR, retinal coloboma (not previously described in the syndrome), corneal opacity, a short lingual frenulum, operated hypertrophic pyloric stenosis, patent foramen ovale, mildly dilated ventricles and a thin corpus callosum; at 14 months she could sit but did not walk or babble [15]. The karyotype showed two cell lines, mos 45,XX,-21[6]/46,XX,del(21)(q22.13q22.3)[24]; FISH confirmed a single copy of DYRK1A in all cells and loss of RUNX1 only in a fraction, and the array delimited a 9.6 Mb terminal deletion with 165 genes, in which DYRK1A was the only one with a ClinGen haploinsufficiency score of 3; the parental karyotypes were normal [15]. The authors interpreted the mosaicism as a post-zygotic error secondary to loss of the centromere of the deleted chromosome and attributed the facial features to DYRK1A haploinsufficiency, which the Face2Gene automated facial analysis ranked first, and not to the mosaic monosomy [15].

4. Clinical Recognition and Warning Signs

4.1 When to suspect

GeneReviews proposes considering the diagnosis in the presence of psychomotor delay or mild to severe ID associated, in infancy or childhood, with any of the following findings: IUGR; microcephaly; typical facial gestalt; neonatal feeding difficulties that may persist; epilepsy (febrile seizures, atonic seizures, absences, generalised myoclonic seizures); hypertonia; abnormal gait; behavioural changes (ASD, anxiety, sleep disturbance); foot anomalies (mild cutaneous syndactyly of toes 2–4, hallux valgus, short fifth toe); visual anomalies (strabismus, myopia, hypermetropia, retinal changes, optic atrophy, coloboma); and urogenital anomalies (cryptorchidism, hypoplastic scrotum, micropenis, inguinal hernia, renal anomalies) [2]. Van Bon and colleagues recommended testing DYRK1A when autism and/or ID are accompanied by microcephaly, especially with two or more of the following: facial gestalt, seizures, hypertonia, feeding problems and abnormal gait [22]. Ji and colleagues suggested including it in the diagnosis of children with Angelman-like features and a distinctive face [8].

4.2 By age and context

In the prenatal period, microcephaly, IUGR, brain or urinary malformations and oligohydramnios may prompt genetic testing, but do not allow the syndrome to be suggested specifically [2, 13]. In early childhood, most children are referred because of the combination of global delay, dysmorphism, microcephaly and feeding difficulties [13]. In the older child, the usual context is ASD with microcephaly and absent or minimal language [9, 22]. In adulthood, diagnosis is rare and arises mainly in the investigation of ID without a previous diagnosis; the case of a 63-year-old woman diagnosed because of multifocal dystonia with stereotypies shows that a late motor picture may be the point of entry, even with normal HC and face [13, 38].

The general approach to microcephaly remains valid: record HC at birth and its evolution to distinguish primary from secondary microcephaly, measure parental HC in borderline forms (−2 to −3 SD), exclude craniosynostosis and look for associated anomalies and neurological signs [29]. In clinical practice, the Oxford Desk Reference applies the term microcephaly to an HC below −3 SD (or the 0.4th centile) and points out that more than 450 syndromes in the London Dysmorphology Database include it [29]. In DYRK1A syndrome, microcephaly may be absent at birth and become evident later, so a normal neonatal measurement does not rule out the diagnosis [13].

4.3 The CS-DYRK1A clinical score

Courraud and colleagues developed a 20-point clinical score (CS-DYRK1A), which the PNDS reproduces and recommends to guide the interpretation of variants [10, 13]. All patients with proven loss-of-function variants had a score of 13 or more (range 13–18.5; mean 15.5), and all patients with other monogenic forms of ID tested (DDX3X, ANKRD11, ARID1B, KMT2A, MED13L, SHANK3 and TCF4) had a score below 13; a score between 10 and 13 is intermediate and below 10 not very suggestive [10]. The version without photographs, with 15 points, discriminates less well [10]. The score also applies to patients initially suspected of having a nucleotide excision repair disorder, whose score did not differ from that of other patients with DYRK1A syndrome [52].

Source: [10, 13].

4.4 Warning and emergency signs

The first seizure, usually febrile, is the most predictable acute event, and the PNDS recommends informing the parents, at the very consultation where the diagnosis is disclosed, about the risk of seizures and the emergency management [13]. Psychomotor regression is not part of the natural history and warrants urgent neurological consultation [13]. Choking, poor weight gain, frequent respiratory infections or food refusal suggest dysphagia with a risk of aspiration and indicate a swallowing study [2]. Recurrent urinary tract infections should prompt investigation for vesicoureteric reflux or a renal anomaly [2, 31]. Reduced vision, leukocoria or new-onset strabismus warrant fundus examination, given the possibility of retinal pathology with a risk of detachment [44, 49].

4.5 Newborn screening

There is no newborn screening and no biochemical or functional marker for the syndrome; non-molecular investigations serve only to assess the extent of the disease and for surveillance [6].

5. Diagnosis: Criteria, Tests and Algorithms

5.1 Diagnostic criterion

The diagnosis is established in a proband with suggestive findings and a heterozygous pathogenic or likely pathogenic DYRK1A variant identified by genetic testing [2, 13]. According to the standards of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP), variants are classified into five categories (pathogenic, likely pathogenic, uncertain significance, likely benign and benign), and the first two are equivalent for clinical decisions [2, 53]. A VUS neither establishes nor excludes the diagnosis [2].

5.2 Tests and strategy

GeneReviews recommends starting with CMA, which detects deletions that include DYRK1A and that sequencing does not detect, and, if this is not diagnostic, moving on to a multigene ID panel or exome sequencing; single-gene testing is rarely useful and is not recommended [2]. The 2010 International Standard Cytogenomic Array (ISCA) consensus established CMA as the first-line cytogenetic test in unexplained developmental delay and ID, ASD and multiple congenital anomalies, with a yield of 15–20%, compared with about 3% for conventional karyotyping [27, 54]. The 2021 ACMG guideline recommends considering exome or genome sequencing as a first- or second-tier test in children with congenital anomalies or ID, because of the higher yield and possible cost reduction when requested early [55]. The French PNDS, for its part, considers next-generation sequencing (small or large ID panels, clinical exome, exome or genome) the preferred strategy, even when suspicion is strong, and stresses that DYRK1A is on all ID panels and that its exons have no capture problems; the array, which detects deletions (about 15% of patients), is performed first-line; Sanger sequencing of the 12 exons has become exceptional; and the karyotype is discussed only when sequencing and array are negative in a highly suggestive phenotype, in search of an apparently balanced translocation [13].

The percentage of variants detectable by a method describes the composition of the spectrum of known variants and should not be confused with the clinical sensitivity of a test in a given patient [2]. A negative result on a panel or exome does not exclude the diagnosis when the phenotype is suggestive. Al-Younis and colleagues described a 7-year-old girl in Canada with global delay, ASD, epilepsy, microcephaly and facial gestalt whose trio exome was inconclusive; the methylation signature was positive for MRD7 and trio genome sequencing identified a de novo deletion of DYRK1A exon 5 (chr21:39 044 532–39 048 467, GRCh38), reclassified from VUS to likely pathogenic [56].

5.3 Variant interpretation

DYRK1A is highly conserved, and in silico predictors tend to overestimate the pathogenicity of missense variants; the CADD (Combined Annotation Dependent Depletion) score discriminates poorly, and conservation beyond vertebrates is more informative [10, 13]. The PNDS recommends functional tests (autophosphorylation or phosphorylation of a target) for missense variants outside the kinase domain, for missense variants within the domain in patients with a low clinical score and for distal truncating variants, and methylation signature analysis for VUS [13]. In the cohort of Courraud and colleagues, the signature of 402 CpG dinucleotides in 165 genes was defined in 10 patients, validated in 6 and compared with 24 and 94 controls; it was positive in six missense variants and in the distal frameshift p.Ser660fs (score 0.92) and negative in p.Arg255Gln, p.Tyr462His and p.Thr588Asn, which were reclassified as likely benign (p.Thr588Asn produced no phenotype in mice) [10]. Truncating variants at amino acids 70–78 of the NM_001396.4 isoform are not pathogenic, and splicing variants affecting only some isoforms call for caution [13, 22]. An apparently missense variant may alter splicing, such as c.1098G>T, which produces an in-frame deletion of 49 amino acids [10].

The report should state the gene, the versioned transcript, the Human Genome Variation Society (HGVS) nomenclature at the complementary DNA (cDNA) and protein levels, zygosity, the ACMG/AMP class and, when the variant is described on the literature transcript and not on MANE, both, to avoid numbering confusion [7, 53]. CNVs are interpreted according to the technical standards of the ACMG and the Clinical Genome Resource (ClinGen) [57]. Parental testing confirms the de novo status (with confirmed paternity and maternity) and, in microdeletions, targeted FISH in the parents looks for balanced insertions, which account for about 2% of apparently de novo interstitial CNVs and which parental CMA, MLPA and quantitative PCR do not detect [2, 11].

5.4 Practical algorithm

5.5 Prenatal and preimplantation diagnosis

When the familial variant is known, prenatal diagnosis (PND) is performed by chorionic villus sampling from 12 weeks of gestation or by amniocentesis from 15 weeks, ideally in the laboratory that studied the index case, weighing the procedure-related risk of fetal loss (0.5–1%) [13]. Preimplantation genetic testing (PGT) is an alternative [2]. The PNDS allows PND to be offered in any pregnancy of the parents of an affected child, although the recurrence risk is below 1% [13]. In the absence of a known familial variant, fetal microcephaly and IUGR may lead to prenatal exome sequencing, although the ultrasound signs do not allow the syndrome to be suspected specifically [13].

6. Differential Diagnosis

6.1 Comparative table

6.2 Main differential diagnoses

Angelman syndrome is the closest differential diagnosis in the young child with microcephaly, absent speech, ataxia and seizures; it is distinguished by the EEG pattern, the facial gestalt and the behaviour, and microcephaly is usually more severe in DYRK1A syndrome, which begins before birth or in the first months, whereas in Angelman syndrome HC is usually below the 25th centile at 3 years and severe microcephaly is rare [2, 13, 59]. Methylation analysis of 15q11-q13 detects about 80% of Angelman cases, and a normal result does not exclude the syndrome, so in an overlapping picture broad genomic tests that include DYRK1A and UBE3A resolve both hypotheses simultaneously [59]. In MECP2-related disorders, the distinguishing element is developmental regression, which is not part of DYRK1A syndrome; the revised 2010 criteria exclude from typical Rett syndrome children with grossly abnormal psychomotor development in the first 6 months, which rules out most infants with DYRK1A syndrome [2, 35, 61]. Microcephalic primordial dwarfisms share IUGR and microcephaly, but the growth deficit is much more severe (HC and height below −4 SD, reaching −8 to −10 SD in MOPD II) and, in MOPD II, the most frequent form, development is relatively preserved; MOPD I combines severe neurological impairment with cortical dysplasia and skeletal anomalies [29, 67]. Mowat-Wilson syndrome combines malformations (Hirschsprung disease, agenesis of the corpus callosum, genitourinary and cardiac anomalies) that are not typical of DYRK1A syndrome, although suspected Hirschsprung disease has been described in a patient with DYRK1A [2, 29, 44].

The overlap with Cockayne syndrome was documented by Le May and colleagues in 11 patients from France, the United Kingdom and Japan initially investigated for a suspected nucleotide excision repair disorder and with a final diagnosis of DYRK1A syndrome; fibroblasts from four of these patients had normal recovery of RNA synthesis and unscheduled DNA synthesis after ultraviolet irradiation and lacked the ATF3 signature, and the authors recommend including DYRK1A in the panels for these disorders [52]. In autosomal recessive primary microcephaly, the absence of extracephalic anomalies and recessive inheritance are the guide; in DYRK1A syndrome, IUGR, feeding difficulties, ocular and urogenital problems and the face differentiate [2]. The more than 450 syndromic pictures with microcephaly and the series of autosomal dominant, recessive and X-linked ID explain why diagnosis rests, in practice, on broad genomic testing [2, 29].

7. Treatments and Risk Management Strategies

7.1 Principles

There is no specific treatment or approved drug for DYRK1A syndrome, and treatment is symptomatic, multidisciplinary and coordinated by a reference team [2, 13]. GeneReviews stated in 2021 that there were no published clinical practice guidelines; the French PNDS (National Diagnosis and Care Protocol), published in October 2021 and prepared by the reference centres for developmental anomalies and malformation syndromes of the South, West and East inter-regions (AnDDI-Rares network), is the only specific national recommendation found and has not been validated by the Haute Autorité de Santé (HAS), which did not take part in its preparation [2, 13]. The sources consulted do not include any subsequent international consensus. The recommendations in this article combine the two sources and point out the divergences.

Early intervention is the cornerstone of treatment: physiotherapy, speech and language therapy and occupational therapy from the first months, with the aim of maximising mobility, communication and autonomy and reducing late orthopaedic complications [2, 13, 35]. The organisation of services differs between countries. GeneReviews describes the United States model (early intervention from 0 to 3 years, developmental preschool from 3 to 5 years with an individualised education plan reviewed annually, school services up to 21 years and a transition plan in adolescence) [2]; the PNDS describes the French model (early medico-social action centres before 6 years and special education and home care services afterwards) and values attendance at nursery in the first years for socialisation [13].

7.2 Assessment after diagnosis

7.3 Feeding and nutrition

Feeding difficulties are almost constant in infants and result from ineffective sucking and swallowing, poor appetite, texture selectivity and GOR [13]. The PNDS recommends early multidisciplinary rehabilitation (speech and language therapy, psychomotor therapy, therapeutic meals, dietetics), a nasogastric tube if nutritional status is insufficient and gastrostomy only rarely, because the difficulties tend to improve in early childhood [13]. GeneReviews accepts gastrostomy if the problem is persistent and recommends a low threshold for clinical swallowing assessment and radiological study in the presence of choking, poor weight gain, recurrent respiratory infections or food refusal, with thickening or chilling of food when useful [2]. GOR is treated with lifestyle and dietary measures and, if necessary, proton pump inhibitors or H2-receptor antagonists, surgery being rare; constipation, sometimes severe, is monitored at each visit [2, 13]. In some patients, the difficulties persist into adulthood and require tube feeding [2].

7.4 Communication and language

Speech and language therapy should begin early and cover feeding and communication, using gestures and sign language, pictures and pictograms, and electronic devices [13]. GeneReviews recommends considering AAC ranging from low technology (picture exchange) to speech-generating devices, and stresses that AAC does not hinder speech development [2]. On the basis of their cohort, Morison and colleagues recommend early access to graphic AAC for all children and therapy targeting apraxia of speech and dysarthria in those who develop speech [41]. Because comprehension exceeds expression, cognitive and educational assessments should use instruments that do not depend on speech [13, 47].

7.5 Motor function, tone and orthopaedics

Early physiotherapy and psychomotor therapy work on tone, sitting position and walking, with a walker in some children and a wheelchair, orthoses and other devices as needed [2, 13]. Occupational therapy addresses fine motor skills, positioning and sensory integration [13]. For hypertonia and dystonia, GeneReviews refers to specialists the decision on baclofen, tizanidine, botulinum toxin (Botox®), antiparkinsonian drugs or orthopaedic surgery, with no specific efficacy data in this syndrome [2]. In the only adult with multifocal dystonia described in detail, levodopa/carbidopa (100/25 mg three times a day for one month) and baclofen (10 mg three times a day for four years) were of no benefit and botulinum toxin was of limited benefit [38]. For scoliosis or foot anomalies, the PNDS recommends annual assessment and follow-up by orthopaedics, which decides on day or night orthoses and possible surgery and guides physiotherapy [13].

7.6 Epilepsy

Treatment of epilepsy is led by an experienced paediatric neurologist or neurologist, with antiseizure medications chosen according to seizure type and tolerability; there is no recommendation of a specific drug and none has been studied specifically in this disorder [2, 13]. The PNDS reports that the efficacy of sodium valproate monotherapy is mentioned in several articles and that epilepsy is usually easily controlled [13]. The review by Okazaki and colleagues found three patients treated with valproate and one with levetiracetam [36]. The case described by these authors also shows the risk of withdrawal: each of the three attempts to withdraw valproate (at 7, 12 and 17 years) was followed by recurrence of afebrile seizures, controlled by reintroduction [36]. There is no syndrome-specific dosing; the valproate and levetiracetam doses in the table in section 7.7 are those of the official labels for focal and generalised seizures, applied by extrapolation [69, 70]. Educating caregivers about safety and about emergency treatment of a prolonged or repeated seizure is part of the care plan [2, 13].

7.7 Sleep

Sleep disturbances, with difficulty falling asleep and frequent waking, are treated according to the cause: adjustment of antiseizure medications, behavioural therapy, sleep hygiene and melatonin [2, 13]. The PNDS indicates melatonin as the pharmacological treatment of these complaints [13]. In the European Union, prolonged-release melatonin Slenyto® (1 mg and 5 mg tablets) was authorised on 20 September 2018 for insomnia in children and adolescents aged 2 to 18 years with ASD and/or Smith-Magenis syndrome. The Commission decision of 26 August 2024 extended the indication to neurogenetic disorders with abnormal daytime melatonin secretion and/or night-time awakenings, and that of 17 March 2025 added insomnia from 6 to 17 years with ADHD, whenever sleep hygiene measures are insufficient. DYRK1A syndrome falls within the 2024 indication by virtue of its neurogenetic nature; the dose has not been studied specifically in this syndrome [71].

None of these drugs has been studied specifically in DYRK1A syndrome, and the doses in the table are those of the labels for the general indications [69–71]. For valproate, the 2018 European measures contraindicate use in girls and women of childbearing potential outside the conditions of the pregnancy prevention programme and require that treatment not be started, even in prepubertal girls, if suitable alternatives exist [72]; in males, the 2024 communication recommends informing about a possible risk of neurodevelopmental disorders in the children of men treated in the 3 months before conception and considering effective contraception, precautionary measures that the Pharmacovigilance Risk Assessment Committee (PRAC) maintained in June 2026 because the evidence is inconsistent [73, 74]. In female adolescents with DYRK1A syndrome and epilepsy, these rules weigh on the choice of antiseizure medication.

7.8 Behaviour and mental health

ASD interventions, including individual applied behaviour analysis (ABA) delivered by a certified analyst, apply to DYRK1A syndrome; the developmental paediatrician guides behavioural strategies or medication (for example, for ADHD) and the child psychiatrist intervenes in severe aggression [2]. The PNDS recommends parental guidance, sibling support and group sessions with a psychologist or educator to develop modes of communication and interaction [13]. The psychiatric heterogeneity (anxiety, depression, ADHD, oppositional behaviour) and the risk of attributing treatable symptoms to ID justify psychiatric and executive function assessment with instruments adapted to minimally verbal people [46, 47].

7.9 Eyes

Strabismus, refractive errors and amblyopia are treated according to usual ophthalmological practice, with optical correction, occlusion and orthoptics; orthoptics may also help with vision-related postural disturbances [13]. Peripheral retinal pathology with avascularity may justify prophylactic laser photocoagulation, as in the cases of Cai and of Lin and colleagues [44, 49]. The frequency of reassessment differs between sources and is detailed in section 11.

7.10 Growth and endocrinology

Growth retardation is assessed with the syndrome-specific charts (Lanvin and colleagues; appendix 2 of the PNDS) and with endocrine investigation when the height curve slows [13, 43]. The PNDS allows growth hormone treatment in proven GH deficiency or in IUGR without catch-up, and levothyroxine in hypothyroidism, with doses and monitoring according to the usual protocols for these indications, which are outside the scope of this article [13]. Lanvin and colleagues propose the same criterion for growth hormone and recommend that a deviation from the specific charts should lead to reassessment of nutritional status and investigation for coeliac disease, chronic inflammatory disease and GH deficiency; at the time of publication, they knew of no patient treated with growth hormone [43]. There are no studies on the efficacy of growth hormone in this syndrome. Growth data in adolescence and adulthood are limited [13].

7.11 Kidney, genital tract, heart and teeth

Renal anomalies and vesicoureteric reflux are followed in paediatric nephrology and urology; cryptorchidism, hypospadias and inguinal hernias are treated surgically according to the usual indications [2, 13]. Congenital heart defects are followed by cardiology according to the lesion [13]. Oral hygiene, with caregiver education, and annual assessment by a dentist or orthodontist prevent the complications of dental calculus and dental anomalies [2, 13]. Eczema may require dermatology [13].

7.12 Transition, adulthood, pregnancy and perioperative care

The PNDS recommends anticipating transition from adolescence: legal protection (guardianship or curatorship in the French system), monitoring of puberty and gynaecological follow-up in women, and screening for common adult comorbidities (cardiovascular risk factors, cancer), whose symptoms may be difficult to identify [13]. The onset of dystonia in the third decade in an isolated case advises continued neurological surveillance in adults [38]. The sources consulted do not describe pregnancies in women with DYRK1A syndrome, and by 2021 no affected people were known to have reproduced [2]. There are no specific anaesthetic recommendations in the sources consulted; preoperative assessment should consider dysphagia and the risk of aspiration, heart disease, epilepsy and communication difficulties.

7.13 Agents to avoid

DYRK1A inhibitors studied for Down syndrome, such as epigallocatechin gallate (EGCG) and Leucettinib-21, act in the opposite direction to what is needed in a haploinsufficiency disorder; in a patient with DYRK1A syndrome, an inhibitor would reduce kinase activity even further and would be expected to be harmful, so these results cannot be extrapolated [6, 38]. In euploid mice, EGCG altered bone parameters, which reinforces caution with supplements containing this compound [21].

7.14 Regulatory status

Source: [2, 13, 18, 71].

7.15 Clinical trials and recent developments

A ClinicalTrials.gov search in October 2026 found no interventional trial in DYRK1A syndrome [18]. The only open study that includes these patients is Simons Searchlight (NCT01238250), an online observational natural history registry of several genetic neurodevelopmental conditions, currently recruiting [18, 45]. Registered trials with DYRK1A modulators concern Down syndrome, where the aim is to reduce the activity of an overexpressed kinase.

Research targeting DYRK1A syndrome is still preclinical. In knock-in mice carrying the human variant p.(Ile48LysfsTer2), chronic lithium treatment started in the neonatal period corrected the brain, behavioural, dendritic, synaptic and phosphoproteomic changes in the juvenile and adult phases [33]. In Dyrk1a mutant mice, an inhibitor of lysine demethylase LSD1 attenuated behavioural deficits, significantly in males in an anxiety test [34]. The description of the interneuron migration defect identifies a circuit target, and patient-derived induced pluripotent stem cell lines, published in 2024 and 2026, allow the disease to be studied in human cells [23, 32]. The 2026 editorial accompanying a themed issue devoted to the syndrome confirms that therapeutic approaches remain experimental [14]. None of these results has clinical translation, and lithium should not be used for this purpose outside trials.

7.16 Divergences between sources

GeneReviews (2021) and the PNDS (2021) agree on the essentials, but diverge on practical points. In testing strategy, GeneReviews places CMA before a panel or exome, whereas the PNDS prefers next-generation sequencing as the main strategy, with the array in parallel for deletions; the 2021 ACMG guideline is closer to the PNDS in considering exome or genome sequencing first- or second-tier tests [2, 13, 55]. On MRI, GeneReviews suggests considering it and the PNDS does not recommend it systematically [2, 13]. On gastrostomy, GeneReviews accepts it for persistent difficulties and the PNDS considers it rare because these improve [2, 13]. For the eyes, reassessment intervals range from three to five years with normal vision (GeneReviews), two to three years without abnormalities and annually with abnormalities (PNDS), and annually by an optometrist until adulthood after 7–8 years (Méjécase and colleagues) [2, 13, 17]. For the frequency of scoliosis, kyphosis and pectus, values range from about 10% to about half, which influences the intensity of orthopaedic surveillance [2, 13].

8. Reproductive Options and Family Counselling

8.1 Inheritance and origin of variants

DYRK1A syndrome is transmitted in an autosomal dominant manner, but almost all cases result from de novo variants [2]. By 2021, all probands whose parents had been tested had a de novo variant [2]. In the cohort of Courraud and colleagues, the variant was de novo in 42 of 50 patients, was inherited from a mosaic father in one case and could not be studied in the parents in seven [10]. In microdeletions, the PNDS points out as an exception the unbalanced translocation inherited from a parent carrying a balanced translocation, and the Oxford Desk Reference recalls that about 2% of apparently de novo interstitial CNVs derive from a balanced insertion in a parent [11, 13].

Parental testing is recommended in all cases [2, 13]. If the variant is not found in the parents' blood and maternity and paternity are confirmed, it is considered de novo; without that confirmation, it should be described as "presumed de novo"; and a negative result in leukocytes does not exclude parental germline or somatic mosaicism [2].

8.2 Recurrence risks by scenario

Recurrence due to parental mosaicism is documented. Agha Gholizadeh and colleagues found, by genome sequencing in Iranian families with several children with ASD, the variant NM_001396.5:c.-77+2T>C (NC_000021.9:g.37365836T>C) in two siblings with microcephaly, motor and speech delay and mild ID, absent in the blood of the parents and of the healthy sibling, and proposed gonadal mosaicism in one of the parents, the first described for this gene [7, 39]. The interpretation has limits: the variant lies in intron 1 of the 5' untranslated region of NM_001396.5, upstream of the start of the MANE Select transcript NM_001347721.2, there was no RNA study or study of mosaicism in sperm, and the two siblings also shared a homozygous 50 bp intronic deletion in DLG2 classified as a VUS [7, 39]. Huang and colleagues, in a Chinese family, found in the fetus of a subsequent pregnancy the same variant NM_001396.5:c.1159C>T p.(Gln387Ter) (NM_001347721.2:c.1132C>T p.(Gln378Ter)) as in the affected sibling, which indicates very probable parental mosaicism [37]. Transmission from a mildly affected parent also occurs: in the Washington cohort, a boy with IQ 133 inherited a pathogenic frameshift, and a carrier parent had IQ 102 and elevated autistic traits [9]. These data justify testing both parents even when they have no evident signs, and explaining that the 1% risk to siblings is an average that does not exclude higher-risk situations.

8.3 Offspring of affected people

By 2021 no people with DYRK1A syndrome were known to have reproduced [2]. The theoretical risk to each child is 50% [2]. In people with a mild phenotype identified in ASD studies, the possibility of reproduction is real and should be addressed in counselling, including the gynaecological follow-up that the PNDS provides for in the transition to adulthood [9, 13].

8.4 Prenatal and preimplantation diagnosis

Counselling before a new pregnancy should discuss the low risk and the options of PND and preimplantation genetic diagnosis, which GeneReviews considers reasonable even with a risk of about 1%, recognising differences in perspective between families and professionals [2]. The PNDS proposes chorionic villus sampling from 12 weeks of gestation or amniocentesis from 15 weeks, with analysis in the laboratory that studied the index case and weighing of the 0.5–1% procedure-related risk of fetal loss [13]. In the Chinese series, amniocentesis was performed at 16+3, 16+5 and 18+5 weeks: one fetus had the sibling's variant and the pregnancy was terminated at the parents' decision, and the other two had normal alleles and the pregnancies continued [37]. Ultrasound does not replace molecular testing: in microcephaly of genetic origin, HC may only decrease at the end of pregnancy (32–36 weeks), and ultrasound-based PND is not reliable [29].

8.5 Counselling of relatives and minors

Cascade testing in relatives other than the parents is not indicated in de novo cases [13]. Healthy siblings of a patient have no indication for predictive testing, unless a parent is a carrier or there are clinical signs. The disclosure consultation should explain the mode of transmission, the usually de novo nature and the absence of parental responsibility, and provide for psychological support [13]. In adulthood, the PNDS recommends a new genetics clinic appointment for counselling of the affected person themselves [13].

9. Special Considerations: Population Diversity and Epidemiology

9.1 Frequency

DYRK1A syndrome accounts for 0.1–0.5% of people with ID and/or autism who have been genetically tested, according to the series of Courcet (2012), O'Roak (2012), the Deciphering Developmental Disorders study (2015) and van Bon (2016) compiled by GeneReviews, and the 2026 editorial places it among the most frequent monogenic causes of ID, with about 0.3–0.5% of cases [2, 14]. These values measure diagnostic yield in selected cohorts, not population prevalence. To give an order of magnitude, ID with IQ below 70 affects 1–2.5% of the population and ASD about 70 per 10 000 people, but multiplying these values does not produce a validated estimate, because the cohorts tested do not represent the population [11, 75]. Orphanet gives a point prevalence below 1/1 000 000, only as a class and based on 54 cases or families recorded in a 2016 source; with more than 30 patients identified in France in 2021 (80 in the French cohort used for the score) and 83 cases in the 2020 literature review cited by Santos Sánchez de las Matas and colleagues, this value reflects the count of described cases and not the real frequency [12, 13, 38]. The sex ratio is about 1 [13].

9.2 Geographical distribution and populations

No founder variants or populations with increased frequency are known; variants are generally private and de novo. The main series come from the United States (University of Washington, Baylor Genetics, Simons Searchlight), Europe (France, Denmark, Switzerland, the Netherlands, the United Kingdom, Italy and Spain), Australia and Asia (China, Japan, South Korea), with cases from Turkey, Iran and Canada [10, 15–17, 31, 36–39, 41, 45, 48, 50, 56]. In Portugal at least one case has been published, diagnosed by exome sequencing in a child followed in Faro in collaboration with the Medical Genetics department of Hospital de Santa Maria, in Lisbon (section 10) [35]. The sources consulted do not include data from Brazil, which does not prove that none exist. The facial gestalt varies with ethnic origin and may be less evident in non-European populations, which reinforces the role of genomic testing and of the clinical score with non-facial components [10, 50].

9.3 Special populations

Adults are underrepresented; the oldest person in the series was 69 years old and the only adult described in neurological detail was 63 years old, with dystonia beginning in the third decade [22, 38]. People with a mild phenotype or without ID are only identified in ASD studies or in family investigation, and their real frequency is unknown [9]. Non-verbal people require adapted cognitive and psychiatric assessment instruments [47].

9.4 Reference networks and patient associations

In Europe, the relevant network is ERN ITHACA, devoted to rare malformation syndromes, ID and other neurodevelopmental disorders [19]. In France, the reference centres for developmental anomalies form part of the AnDDI-Rares network and collaborate with the DéfiScience network [13]. GeneReviews mentions the specialised DYRK1A clinic at Radboudumc, in Nijmegen, the Netherlands [2]. The DYRK1A Syndrome International Association (DSIA), founded in 2017, brings families together and provides information at www.dyrk1a.org; there is also a family group, "Dyrk1a Awareness & Information", on a social network [13, 20]. The Simons Searchlight registry collects natural history data with remote participation [18].

10. Illustrative Clinical Case

10.1 Presentation

Oliveira and colleagues, from the Centro Hospitalar Universitário do Algarve (Faro) and the Medical Genetics Department of Hospital de Santa Maria (Lisbon), described a girl followed in a developmental clinic from 9 months of age for psychomotor delay [35]. She was born after an uneventful term pregnancy, by vaginal delivery; newborn screening was normal and the family history was non-contributory [35]. On the first day of life, examination showed irritability and a wide anterior fontanelle, and transfontanellar ultrasound revealed mild dilatation of the lateral and third ventricles, confirmed by MRI at 5 months (Figure 1, https://pmc.ncbi.nlm.nih.gov/articles/PMC10829433/figure/FIG1/) [35]. HC remained below the 3rd centile until 5 months and then at the 3rd centile [35].

10.2 Course and findings

In the first years, developmental milestones were globally delayed, with expressive language most affected; neurological examination showed axial hypotonia with appendicular hypertonia [35]. The child had periods of irritability, sleep disturbance and stereotypies; swallowing difficulties appeared with the introduction of solids at 6 months; and she had frequent urinary tract infections in the first year without underlying urogenital malformation [35]. During follow-up, deep-set eyes, slightly upslanting palpebral fissures, a high nasal bridge with a broad nasal tip, a thin upper lip, widely spaced teeth, retrognathia and mild pectus excavatum were recorded (Figures 2 and 3, https://pmc.ncbi.nlm.nih.gov/articles/PMC10829433/figure/FIG2/ and https://pmc.ncbi.nlm.nih.gov/articles/PMC10829433/figure/FIG3/) [35].

10.3 Diagnosis

The first genetics clinic appointment took place at 3 years. Array comparative genomic hybridisation showed no CNVs, and exome sequencing identified the heterozygous variant NM_001396.5:c.894del p.(Lys299AsnfsTer7), classified as likely pathogenic [35]. On the MANE transcript, the same variant corresponds to NM_001347721.2:c.867del, p.(Lys290AsnfsTer7) (NC_000021.9:g.37490404del) [7]. The diagnostic pathway matches the sequence recommended by GeneReviews (array followed by exome) and shows the usual delay between the first signs and the molecular diagnosis [2].

10.4 Management and course

The child was assessed with regard to development, feeding and the urogenital, cardiovascular, ocular and neurological systems, and received genetic counselling [35]. She started physiotherapy, speech and language therapy and occupational therapy at 9 months, which she was still receiving at 7 years [35]. She began to walk unaided at 20 months and said her first words at 17 months; at 7 years she constructed three-word sentences, had increasing autonomy in hygiene and feeding and was progressing at school, with concentration difficulties [35].

10.5 Commentary

The case brings together several elements of the CS-DYRK1A score: expressive language delay, microcephaly, feeding difficulties, ventricular dilatation, limb hypertonia, stereotypies and a face with enophthalmos, a thin upper lip, widely spaced teeth and retrognathia [13, 35]. The course is more favourable than the cohort average, in which walking usually appears after 2 years and most are left with minimal language, which illustrates the variability of expression and the difficulty of predicting individual prognosis from the variant type [2, 16, 41]. The absence of seizures and of a formal diagnosis of ASD by 7 years did not rule out the diagnosis. The report does not describe the results of the echocardiogram, renal ultrasound or ophthalmological examination, which are recommended in all patients [2, 13].

11. Prognosis and Long-Term Follow-up

11.1 Prognosis

Life expectancy does not appear to be reduced, and adults up to 69 years have been described, but longitudinal data on behavioural and neurological evolution are limited [2, 22]. ID and language impairment are permanent, and most people are left with moderate to severe ID and minimal verbal language, needing AAC [2, 41]. Microcephaly worsens in standard deviation terms throughout childhood [2]. Feeding difficulties tend to improve in early childhood, although they persist in some into adulthood [2, 13]. Epilepsy is usually controlled, but may recur when treatment is withdrawn [13, 36]. In adulthood, the onset of progressive multifocal dystonia has been described, and adults should be monitored for new motor manifestations and for the common comorbidities of age [13, 38]. Ocular complications, especially retinal detachment, may compromise vision if not detected [17, 44].

11.2 Organisation of follow-up

The PNDS recommends annual follow-up by the family doctor or paediatrician and by the paediatric neurologist or neurologist, and assessment by the reference centre every two years, maintaining this frequency after 18 years with a new genetic counselling appointment [13]. At each visit, nutritional status and growth, epileptic phenomena, any regression (which requires urgent neurological consultation), the spine and feet, teeth, eyes, behaviour and sleep are assessed [13]. GeneReviews recommends lifelong follow-up by specialists for any cardiac, ocular and dental problems present [2]. In groups with undiagnosed ID, the Oxford Desk Reference suggests a review one year and another two years after the initial consultation; once the diagnosis is established, the frequency is that of the following table [11].

11.3 Surveillance table

12. Summary and Final Considerations

DYRK1A syndrome results from haploinsufficiency of a highly conserved kinase at 21q22.13, essential for brain growth and organisation, and lies at the opposite end from Down syndrome on the same gene-dosage axis. The picture is recognisable: microcephaly of prenatal or early onset that progresses, ID that is usually moderate to severe, minimal verbal language with relatively preserved comprehension, early feeding difficulties, febrile seizures and generally controllable epilepsy, autistic traits and anxiety, ataxic gait with limb hypertonia, a face with deep-set eyes and retrognathia, and ocular, renal, genital, cardiac and skeletal anomalies that should be sought systematically. The 21q22.13 microdeletion forms have the same phenotype, unless the deletion includes other genes.

Diagnosis rests on broad genomic testing, with the array and panel, exome or genome sequencing, and with the methylation signature and functional assays for variants that are difficult to interpret. The existence of two reference transcripts, the 754-amino-acid MANE transcript and the 763-amino-acid literature transcript, makes it necessary to state both in the report. The CS-DYRK1A score helps to decide when a variant is compatible with the syndrome, and variable expressivity, including people without ID and transmission from mildly affected parents, calls for caution in counselling.

Treatment is supportive and multidisciplinary, centred on early intervention, AAC, feeding, epilepsy, sleep and behaviour, with ophthalmological, renal, cardiac, dental and orthopaedic surveillance according to the 2021 French PNDS, the only specific recommendation, and GeneReviews. There are no approved drugs and no clinical trials in the syndrome; DYRK1A inhibitors tested in Down syndrome act in the opposite direction to what is needed, and the approaches that look promising in animal models, such as neonatal lithium or LSD1 inhibition, remain preclinical. The recurrence risk for siblings is low, of the order of 1%, but documented parental mosaicism justifies testing the parents and discussing the options of PND and preimplantation diagnosis. Population frequency studies, longitudinal data in adulthood and data from Portugal and Brazil beyond isolated cases are lacking.

13. References

1. OMIM. #614104 Intellectual developmental disorder, autosomal dominant 7; MRD7. Johns Hopkins University; last edited 2022-04-01. Available from: https://omim.org/entry/614104
2. van Bon BWM, Coe BP, de Vries BBA, Eichler EE. DYRK1A Syndrome. 2015 Dec 17 [updated 2021 Mar 18]. In: Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK333438/
3. Orphanet. Intellectual disability syndrome due to a DYRK1A point mutation (ORPHA:464311). Available from: https://www.orpha.net/en/disease/detail/464311
4. Orphanet. DYRK1A-related intellectual disability syndrome due to 21q22.13q22.2 microdeletion (ORPHA:268261). Available from: https://www.orpha.net/en/disease/detail/268261
5. OMIM. *600855 Dual-specificity tyrosine phosphorylation-regulated kinase 1A; DYRK1A. Johns Hopkins University. Available from: https://omim.org/entry/600855
6. GeneGuidance. DYRK1A — clinical curation of the gene and the MRD7 condition (OMIM 614104). GeneGuidance database (version of 2026-09-26).
7. VariantValidator. Normalisation of DYRK1A variants on GRCh38 (NM_001396.5, NM_001347721.2 MANE Select). Available from: https://rest.variantvalidator.org/ [accessed 31 Oct 2026].
8. Ji J, Lee H, Argiropoulos B, Dorrani N, Mann J, Martinez-Agosto JA, et al. DYRK1A haploinsufficiency causes a new recognizable syndrome with microcephaly, intellectual disability, speech impairment, and distinct facies. Eur J Hum Genet. 2015;23(11):1473-81. doi:10.1038/ejhg.2015.71. PMID: 25944381; PMC4613469.
9. Kurtz-Nelson EC, Rea HM, Petriceks AC, Hudac CM, Wang T, Earl RK, et al. Characterizing the autism spectrum phenotype in DYRK1A-related syndrome. Autism Res. 2023;16(8):1488-1500. doi:10.1002/aur.2995. PMID: 37497568; PMC10530559.
10. Courraud J, Chater-Diehl E, Durand B, Vincent M, Del Mar Muniz Moreno M, Boujelbene I, et al. Integrative approach to interpret DYRK1A variants, leading to a frequent neurodevelopmental disorder. Genet Med. 2021;23(11):2150-2159. doi:10.1038/s41436-021-01263-1. PMID: 34345024.
11. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Intellectual disability. p. 188–193.
12. Orphanet. DYRK1A-related intellectual disability syndrome (ORPHA:464306). Available from: https://www.orpha.net/en/disease/detail/464306
13. Willems M, Vincent M, Piton A, et al.; Centres de référence Anomalies du développement et syndromes malformatifs (filière AnDDI-Rares). Protocole national de diagnostic et de soins (PNDS): Syndrome DYRK1A. Paris: Haute Autorité de Santé; 2021 [published 14 Oct 2021]. Available from: https://www.has-sante.fr/jcms/p_3290097/fr/syndrome-dyrk1a
14. Glass OK, Piton A, Pfalzer A. Editorial: Advances in DYRK1A syndrome: underlying mechanisms, disease models, and novel therapeutic approaches. Front Neurosci. 2026;20:1803346. doi:10.3389/fnins.2026.1803346. PMID: 41799883; PMC12960572.
15. Taşdelen E, Tekbaş UC, Baş H, Sezer A. Two remarkable cases of haploinsufficiency found in the DYRK1A gene. Acta Neurol Belg. 2025;125(3):839-843. doi:10.1007/s13760-024-02685-9. PMID: 39572511.
16. Earl RK, Turner TN, Mefford HC, Hudac CM, Gerdts J, Eichler EE, et al. Clinical phenotype of ASD-associated DYRK1A haploinsufficiency. Mol Autism. 2017;8:54. doi:10.1186/s13229-017-0173-5. PMID: 29034068; PMC5629761.
17. Méjécase C, Way CM, Owen N, Moosajee M. Ocular Phenotype Associated with DYRK1A Variants. Genes (Basel). 2021;12(2):234. doi:10.3390/genes12020234. PMID: 33562844; PMC7915179.
18. ClinicalTrials.gov. Search "DYRK1A" (API v2). National Library of Medicine. Available from: https://clinicaltrials.gov/ [accessed 31 Oct 2026].
19. ERN ITHACA — European Reference Network on Rare Malformation Syndromes, Intellectual and Other Neurodevelopmental Disorders. Available from: https://ern-ithaca.eu/ [accessed 31 Oct 2026].
20. DYRK1A Syndrome International Association. Available from: https://www.dyrk1a.org/ [accessed 31 Oct 2026].
21. Otte ED, Roper RJ. Skeletal health in DYRK1A syndrome. Front Neurosci. 2024;18:1462893. doi:10.3389/fnins.2024.1462893. PMID: 39308945; PMC11413744.
22. van Bon BW, Coe BP, Bernier R, Green C, Gerdts J, Witherspoon K, et al. Disruptive de novo mutations of DYRK1A lead to a syndromic form of autism and ID. Mol Psychiatry. 2016;21(1):126-32. doi:10.1038/mp.2015.5. PMID: 25707398; PMC4547916.
23. Plećaš A, Gough G, Kunštek S, Babić M, Murray A, Nižetić D, et al. DYRK1A - a gene dose paradox. Front Cell Dev Biol. 2026;14:1963992. doi:10.3389/fcell.2026.1963992. PMID: 42909264; PMC13683934.
24. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Down's syndrome (trisomy 21). p. 666–669.
25. Møller RS, Kübart S, Hoeltzenbein M, Heye B, Vogel I, Hansen CP, et al. Truncation of the Down syndrome candidate gene DYRK1A in two unrelated patients with microcephaly. Am J Hum Genet. 2008;82(5):1165-70. doi:10.1016/j.ajhg.2008.03.001. PMID: 18405873; PMC2427221.
26. Bronicki LM, Redin C, Drunat S, Piton A, Lyons M, Passemard S, et al. Ten new cases further delineate the syndromic intellectual disability phenotype caused by mutations in DYRK1A. Eur J Hum Genet. 2015;23(11):1482-7. doi:10.1038/ejhg.2015.29. PMID: 25920557; PMC4613470.
27. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Deletions and duplications (including microdeletions and microduplications). p. 660–665.
28. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Genetics, genetic counseling, and prevention — chromosomal/genomic imbalance. p. 929–937.
29. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Microcephaly. p. 228–231.
30. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Autism and autism spectrum disorders. p. 350–353.
31. Blackburn ATM, Bekheirnia N, Uma VC, Corkins ME, Xu Y, Rosenfeld JA, et al. DYRK1A-related intellectual disability: a syndrome associated with congenital anomalies of the kidney and urinary tract. Genet Med. 2019;21(12):2755-2764. doi:10.1038/s41436-019-0576-0. PMID: 31263215; PMC6895419.
32. Hinckelmann MV, Dubos A, Artot V, Rudolf G, Nguyen TL, Tilly P, et al. Interneuron migration defects during corticogenesis contribute to Dyrk1a haploinsufficiency syndrome pathogenesis. Mol Psychiatry. 2025;30(11):5227-5244. doi:10.1038/s41380-025-03109-7. PMID: 40634533.
33. Roh JD, Bae M, Kim H, Yang Y, Lee Y, Cho Y, et al. Lithium normalizes ASD-related neuronal, synaptic, and behavioral phenotypes in DYRK1A-knockin mice. Mol Psychiatry. 2025;30(6):2584-2596. doi:10.1038/s41380-024-02865-2. PMID: 39633007; PMC12092274.
34. Lin CH, Yu M, Patel PJ, Li P, Yan Z. Epigenetic treatment of synaptic and behavioral deficits in Dyrk1a-mutant mice. Neuropsychopharmacology. 2026;51(8):1413-1423. doi:10.1038/s41386-026-02353-x. PMID: 41559452; PMC13291241.
35. Oliveira Í, Fernandes A, Pereira M, Rodrigues M, Silva N, Mendonça C. A Rare Cause of Intellectual Disability. Cureus. 2024;16(1):e51451. doi:10.7759/cureus.51451. PMID: 38298296; PMC10829433.
36. Okazaki T, Yamada H, Matsuura K, Kasagi N, Miyake N, Matsumoto N, et al. Clinical course of epilepsy and white matter abnormality linked to a novel DYRK1A variant. Hum Genome Var. 2021;8(1):26. doi:10.1038/s41439-021-00157-7. PMID: 34253714; PMC8275604.
37. Huang C, Luo H, Zeng B, Feng C, Chen J, Yuan H, et al. Identification of two novel and one rare mutation in DYRK1A and prenatal diagnoses in three Chinese families with intellectual Disability-7. Front Genet. 2023;14:1290949. doi:10.3389/fgene.2023.1290949. PMID: 38179410; PMC10765505.
38. Santos Sánchez de Las Matas L, Ordás C, Almoguera B, Grondona FL, Coma AQ. Complex Multifocal Dystonia, Intellectual Disability, and Stereotypies Due to a Novel DYRK1A Variant. Mov Disord Clin Pract. 2025;12(7):1011-1013. doi:10.1002/mdc3.70013. PMID: 39969062; PMC12275002.
39. Agha Gholizadeh M, Behjati F, Ghasemi Firouzabadi S, Heidari E, Razmara E, Almadani N, et al. Novel splicing variant and gonadal mosaicism in DYRK1A gene identified by whole-genome sequencing in multiplex autism spectrum disorder families. Neurogenetics. 2024;25(4):377-391. doi:10.1007/s10048-024-00768-6. PMID: 38976082.
40. Boßelmann CM, Leu C, Brünger T, Hoffmann L, Baldassari S, Chipaux M, et al. Analysis of 1386 epileptogenic brain lesions reveals association with DYRK1A and EGFR. Nat Commun. 2024;15(1):10429. doi:10.1038/s41467-024-54911-w. PMID: 39616148; PMC11608322.
41. Morison LD, Braden RO, Amor DJ, Brignell A, van Bon BWM, Morgan AT. Social motivation a relative strength in DYRK1A syndrome on a background of significant speech and language impairments. Eur J Hum Genet. 2022;30(7):800-811. doi:10.1038/s41431-022-01079-w. PMID: 35437318; PMC9259653.
42. Durand B, Schaefer E, Burger P, Baer S, Schroder C, Mandel JL, et al. Neurocognitive and neurobehavioral characterization of two frequent forms of neurodevelopmental disorders: the DYRK1A and the Wiedemann-Steiner syndromes. Clin Genet. 2022;102(4):296-304. doi:10.1111/cge.14190. PMID: 35821609.
43. Lanvin PL, Goronflot T, Isidor B, Nizon M, Durand B, El Chehadeh S, et al. Growth charts in DYRK1A syndrome. Am J Med Genet A. 2024;194(1):9-16. doi:10.1002/ajmg.a.63412. PMID: 37740550.
44. Lin S, Hay E, Thompson DA, Moosajee M, Webster AR, Mahroo OA, et al. DYRK1A syndrome presenting with a familial exudative vitreoretinopathy (FEVR)-like retinovascular phenotype. Ophthalmic Genet. 2025;46(4):389-393. doi:10.1080/13816810.2025.2503388. PMID: 40405340; PMC12312742.
45. Fenster R, Ziegler A, Kentros C, Geltzeiler A, Green Snyder L, Brooks E, et al. Characterization of phenotypic range in DYRK1A haploinsufficiency syndrome using standardized behavioral measures. Am J Med Genet A. 2022;188(7):1954-1963. doi:10.1002/ajmg.a.62721. PMID: 35285131.
46. Neuhaus E, Rea H, Jones E, Benavidez H, Miles C, Whiting A, et al. Shared and divergent mental health characteristics of ADNP-, CHD8- and DYRK1A-related neurodevelopmental conditions. J Neurodev Disord. 2024;16(1):15. doi:10.1186/s11689-024-09532-1. PMID: 38622540; PMC11017562.
47. Rea HM, Webb SJ, Kurtz-Nelson EC, Hudac CM, Bernier RA, Miles C, et al. Characterizing executive functioning and associated behaviors in individuals with dual-specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A) syndrome. Front Neurosci. 2025;18:1485499. doi:10.3389/fnins.2024.1485499. PMID: 39840013; PMC11747226.
48. Yang SA, Boo JH, Choi H, Kim SJ, Cheon CK, Lee SM, et al. Ocular manifestations in DYRK1A mutation: a five-case series from South Korea. Ophthalmic Genet. 2026;47(4):420-424. doi:10.1080/13816810.2026.2635639. PMID: 41804029.
49. Cai L, Wakabayashi T, Yonekawa Y, Wasserman BN. DYRK1A retinopathy. J AAPOS. 2023;27(2):107-110. doi:10.1016/j.jaapos.2022.12.002. PMID: 36736451.
50. Moroni A, Pezzani L, Alfei E, Scatigno A, Cereda A, Marzaroli M, et al. Rock around DYRK1A: Ethnic diversity, clinical challenges. Am J Med Genet A. 2023;191(5):1459-1464. doi:10.1002/ajmg.a.63140. PMID: 36772973.
51. Infantino I, Tocchioni F, Ghionzoli M, Coletta R, Morini F, Morabito A. Case Report: Gut and spleen anomalies associated with DYRK1A syndrome. Front Pediatr. 2022;10:936732. doi:10.3389/fped.2022.936732. PMID: 36741085; PMC9890171.
52. Le May N, Courraud J, Boujelbène I, Obringer C, Ogi T, Lehmann AR, et al. Clinical and molecular overlap between nucleotide excision repair (NER) disorders and DYRK1A haploinsufficiency syndrome. Front Neurosci. 2025;19:1554093. doi:10.3389/fnins.2025.1554093. PMID: 40206408; PMC11979163.
53. 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; PMC4544753.
54. 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; PMC2869000.
55. 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.
56. Al-Younis I, Basque L, Crapoulet N, Dyack S, Del Caño-Ochoa F, Ramón-Maiques S, et al. Episignature leads to diagnosis and reclassification of DYRK1A variant in a child with syndromic neurodevelopmental disorder: a case report. Front Genet. 2026;17:1813300. doi:10.3389/fgene.2026.1813300. PMID: 42256340; PMC13236001.
57. 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):245-257. doi:10.1038/s41436-019-0686-8. PMID: 31690835; PMC7313390.
58. OMIM. #105830 Angelman syndrome; AS. Available from: https://omim.org/entry/105830
59. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Angelman syndrome. p. 346–349.
60. OMIM. #312750 Rett syndrome; RTT. Available from: https://omim.org/entry/312750
61. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Rett syndrome. p. 520–523.
62. OMIM. #610954 Pitt-Hopkins syndrome; PTHS. Available from: https://omim.org/entry/610954
63. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Pitt-Hopkins syndrome. p. 272–273.
64. OMIM. #235730 Mowat-Wilson syndrome; MOWS. Available from: https://omim.org/entry/235730
65. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. 1p36 deletion syndrome. p. 76–79.
66. OMIM. #617140 ZTTK syndrome; ZTTKS. Available from: https://omim.org/entry/617140
67. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Microcephalic primordial dwarfing syndromes (MOPD II, Seckel syndrome, MOPD I). p. 128–131.
68. OMIM. #300749 Intellectual developmental disorder and microcephaly with pontine and cerebellar hypoplasia; MICPCH. Available from: https://omim.org/entry/300749
69. AbbVie. DEPAKOTE (divalproex sodium) delayed-release tablets — prescribing information. NDA 018723. Effective revision 2026-03-31. DailyMed set id 08a65cf4-7749-4ceb-6895-8f4805e2b01f. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=08a65cf4-7749-4ceb-6895-8f4805e2b01f
70. UCB, Inc. KEPPRA (levetiracetam) tablets, oral solution — prescribing information. NDA 021035/021505. Effective revision 2026-01-28. DailyMed set id 3ca9df05-a506-4ec8-a4fe-320f1219ab21. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=3ca9df05-a506-4ec8-a4fe-320f1219ab21
71. European Medicines Agency. Slenyto 1 mg and 5 mg prolonged-release tablets: summary of product characteristics. First authorisation 20 Sep 2018 (ASD and/or Smith-Magenis syndrome); renewal 5 Jun 2023; extension to neurogenetic disorders (II/0025) 26 Aug 2024; ADHD indication (II/0028) 17 Mar 2025; product information updated on 10 Jun 2026 [accessed 7 Nov 2026]. Amsterdam: EMA. Available from: https://www.ema.europa.eu/en/documents/product-information/slenyto-epar-product-information_en.pdf
72. European Medicines Agency. Valproate and related substances – Article 31 referral: new measures to avoid valproate exposure in pregnancy endorsed. EMA/375438/2018. Amsterdam: EMA; 2018. Available from: https://www.ema.europa.eu/en/medicines/human/referrals/valproate-related-substances-0
73. European Medicines Agency. Direct healthcare professional communication (DHPC): Valproate-containing medicines: new measures regarding the potential risk of neurodevelopmental disorders in children of fathers treated with valproate in the 3 months prior to conception. Amsterdam: EMA; 2024. Available from: https://www.ema.europa.eu/en/documents/dhpc/direct-healthcare-professional-communication-dhpc-valproate-containing-medicines-new-measures-regarding-potential-risk-neurodevelopmental-disorders-children-fathers-treated-valproate-3-months-prior_en.pdf
74. Pharmacovigilance Risk Assessment Committee (PRAC). PRAC recommendations on signals adopted at the 8-11 June 2026 PRAC meeting. EMA/PRAC/124080/2026. Amsterdam: EMA; 2026. Section 1.3, Valproate and related substances – neurodevelopmental disorders with paternal exposure (EPITT 20191). Available from: https://www.ema.europa.eu/en/documents/prac-recommendation/prac-recommendations-signals-adopted-8-11-june-2026-prac-meeting_en.pdf
75. Loiola JCR, Alexandre AF, Prata VPS, Ribeiro EM. Autismo ou transtorno do espectro autista (TEA). In: Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 142–145.